ORNL
&EPA
Oak Ridge
National
Laboratory
ORNL/BS-197
United States
Environmental Protection
Agency
Office of Health and
Environmental Assessment
Washington, DC 20460
EPA-600/9-82-001
February 1982
Research and Development
Assessment of Risks to
Human Reproduction and to
Development of the Human
Conceptus from Exposure to
Environmental Substances
Proceedings of U.S. Environmental Protection Agency-
Sponsored Conferences:
October 1-3, 198O, Atlanta, Georgia, and
December 7-10, 1980, St. Louis, Missouri
-------
Printed in the United States of America. Available from
National Technical Information Service
U.S. Department of Commerce
5285 Port Royal Road, Springfield, Virginia 22161
NTIS price codes—Printed Copy: A09 Microfiche A01
This report was prepared as an account of work sponsored by an agency of the
United StatesGovernment, Neither the U nited States Government nor any agency
thereof, nor any of their employees, makes any warranty, express or implied, or
assumes any legal liability or responsibility for the accuracy, completeness, or
usefulness of any information, apparatus, product, or process disclosed, or
represents that its use would not infringe privately owned rights. Reference herein
to any specific commercial product, process, or service by trade name, trademark,
manufacturer, or otherwise, does not necessarily constitute or imply its
endorsement, recommendation, or favoring by the United States Government or
any agency thereof. The views and opinions of authors expressed herein do not
necessarily stale or reflect those of the United StatesGovernment or any agency
thereof.
-------
ORNLyEIS-197
EPA-600/9-82-001
Contract No. W-7405-eng-26
ASSESSMENT OF RISKS TO HUMAN REPRODUCTION
AND TO DEVELOPMENT OF THE HUMAN CONCEPTUS
FROM EXPOSURE TO ENVIRONMENTAL SUBSTANCES
Proceedings of U.S. Environmental Protection Agency-
Sponsored Conferences:
October 1-3, 1980, Atlanta, Georgia,
and
December 7-10, 1980, St. Louis, Missouri
Project Officers:
Wayne M. Galbraith, Ph.D.
Peter Voytek, Ph.D.
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
and
Michael G. Ryon, M.S.
Chemical Effects Information Center
Information Center Complex
Information Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37830
Work sponsored by the U.S. Environmental Protection Agency,
Washington, D.C., under Interagency Agreements
No. 80-D-X1011 and No. 81-D-X0453
Published: February 1982
OAK RIDGE NATIONAL LABORATORY
Oak Ridge, Tennessee 37830
operated by
UNION CARBIDE CORPORATION
for the
DEPARTMENT OF ENERGY
-------
-------
IX
1
CONTENTS
Tables
Acknowledgments
CHAPTER 1
Introduction
CHAPTER 2
Female Reproduction 3
General Reproductive Toxicity Screen 3
Qualitative Reproductive Toxicity Tests 5
Estrogen agonist-antagonist 5
Androgen agonist-antagonist 5
Nonsteroidal toxicant screening tests 6
Computerized integrated data base 6
Quantitative Reproductive Toxicity Tests 6
Risk Assessment 7
Research Needed 9
Qualitative reproductive toxicity tests 9
Quantitative reproductive toxicity tests 10
Specific recommendations 10
Extrapolation of animal data to humans 11
References 12
Appendix 13
Details of Test Protocols and Glossary of Terms for
Female Risk Assessments 13
Description and Discussion of Tests Useful in Assessing
Risk to the Female Reproductive System 13
Qualitative Reproductive Toxicity Tests 13
Estrogen agonist-antagonist 13
Androgen agonist-antagonist 14
Nonsteroidal toxicant screening tests 14
Quantitative Reproductive Toxicity Tests 18
Estrogen agonist-antagonist 18
Androgen agonist-antagonist 19
Hypothalamic-Pituitary Function Tests 21
Assay of agents that stimulate the release of
gonadotropins from cells of the anterior
pituitary gland 21
in
-------
Assays of agents that inhibit the release of
gonadotropins from cells of the anterior
pituitary gland 22
Assay of an agent that inhibits the release of
prolactin from cells of the anterior pituitary
gland 23
Assay of an agent that inhibits the release of
prolactin from pituitary cells 23
Assay of the activity of an agent that alters the
secretion of dopamine by hypothalamic
neurons 24
Assay of the activity of an agent that alters the
secretion of norepinephrine by hypothalamic
neurons 24
Assay of the activity of an agent that alters the
secretion of GnRH . • • 24
Assay of the activity of an agent that alters the
secretion of hypothalamic opioid peptides 25
Blood flow of the hypothalamic-hypophysial system . . 25
Sexual behavior tests 25
Ovarian Toxicity 27
Oocyte and follicle toxicity 27
Inhibition of steroidogenesis 28
References 34
Glossary of Terms Used in Female Reproduction 37
CHAPTERS
Considerations in Evaluating Risk to Male
Reproduction 41
Introduction 41
Aspects of the Problem 42
Selection of an Animal Model 43
Tests for Evaluating Reproductive Damage 46
Evaluation of Reproductive Damage in Exposed or
Potentially Exposed Men 52
General 52
Surveillance studies 52
Study of men with known toxic exposure 53
Additional comment on human testing procedures . . .54
Assessment of risk to men 55
Protocols for Testing Compounds with Animal Models . . 57
Test 1 — initial screen 57
Test 2 — dose response curve 57
Test 3 — recovery study 59
iv
-------
Research Needed 60
References '. 63
Appendix 69
Details of Test Protocols and Glossary of Terms for
Male Risk Assessment 69
Description and Discussion of Tests Useful in
Animal Models or Man 69
Body weight 69
Testicular Characteristics 69
Testis size in situ 69
Testis weight 70
Spermatid reserves 70
Histopathological analysis of testes 71
Counts of preleptotene or leptotene
spermatocytes 72
Epididymal Characteristics 72
Weight of distal half of epididymis 72
Number of sperm in the distal half of
epididymis 72
Motility of sperm from the distal end 72
Gross morphology of spermatozoa from the
distal end 73
Detailed morphology of spermatozoa from the
distal end 73
Accessory Sex Gland Characteristics 73
Seminal Analysis 73
General aspects of seminal analysis 73
Volume 75
Seminal plasma constituents 75
Spermatozoa! concentration 76
Total sperm per ejaculate 76
Sperm motility 77
Spermatozoal morphology 78
Ejaculated sperm as an in vitro test system 79
Assessment of Male Reproductive Toxicity Using
Endocrinological Methods 79
.General 79
Hormone assay and application 80
Examination of Known Toxic Exposures 82
Humans 82
Animal models 83
-------
Fertility Testing .84
Tests available 84
Usefulness 85
Sensitivity 85
Specificity 86
Sperm Nucleus Integrity 86
Quinacrine staining for Y-chromosome
aneuploidy 86
Spermatozoal morphology 87
Karyotyping of human spermatozoa by the
denuded-hamster-egg technique 87
Dose Response 87
References 88
Glossary of Terms Used in Male Reproduction 93
CHAPTER 4
Current Status of, and Considerations for, Estimation
of Risk to the Human Conceptus from
Environmental Chemicals 99
Definition and Scope 99
Impact of Developmental Abnormalities on Humans ... 99
Causes of Congenital Malformations 100
Qualitative Evaluation of Risk Potential 100
Interspecies comparisons 100
Dosing and mode of administration 101
Placenta! transfer 102
Pharmacokinetics and metabolism 103
Mechanisms of action 103
Animal Studies 104
Standard teratogenicity testing 104
Functional teratogenicity testing 106
Short-Term Testing Procedures 107
Prioritizing of chemicals for in-depth study 107
Characteristics of short-term assays 108
Potential short-term systems 109
Quantitative Risk Assessment 110
Priorities for Future Research in Teratology 112
References 113
CHAPTERS
Other Considerations: Epidemiology, Pharmacokinetics,
and Sexual Behavior 117
Epidemiology: Methods and Limitations 117
Hypothesis generating studies 117
VI
-------
Analytic studies for formally testing hypotheses
and quantifying risks 118
Limitations 118
Possible data sources and useful approaches 120
Pharmacokinetics . „ . 121
Sexual Behavior 123
Introduction 123
Qualitative evaluation of risk potential 128
Animal studies 131
Assessment of human sexual behavior:
surveillance and epidemiological
studies 136
Priority areas for future research 139
References 141
Steering Committee 145
Participants ' 147
Reviewers 151
Index 153
VII
-------
-------
TABLES
1. Reproductive Processes Potentially Susceptible
to Reproductive Toxicants 8
2. Estrogen Agonist Screen 14
3. Time Table for Intergenerational Protocol to Evaluate
Putative Toxicant Effects on Reproduction in Sexually
Mature Animals 15
4. Compounds Tested for Oocyte/Follicle Toxicity in
the Murine Assay 29
5. Agents That Inhibit Steroidogenesis 31
6. Features of Ovarian Cell Preparations, In Vitro,
Potentially Useful in Xenobiotic Inhibition of
Steroidogenesis 33
7. Criteria for Evaluation of Male Reproduction in
Favored Animal Models and Man 44
8. Te,sts Considered Useful for Screening Toxic
Compounds 47
9. Reasons for Rejection of Potential Evaluation
Tests Considered by Male Reproductive
Subgroup 49
10. Approximate Variation Between Animals for
Suggested Test Criteria (CV) Coefficient of
Variation (%) 50
11. Chronology of Conduct for Test with Animal
Models 58
12. Some In Vitro Short-Term Systems Currently
in Various Stages of Development 110
IX
-------
-------
ACKNOWLEDGMENTS
This document resulted from discussions at two conference
meetings sponsored by the U.S. Environmental Protection Agency
(EPA). The first meeting took place in Atlanta, Georgia, October
1-3, 1980, and the second, in St. Louis, Missouri, December 7-10,
1980. The participants in this project are listed at the back of this
document; their interest, scientific knowledge, and contributions of
personal and professional time are largely responsible for the
production of this report. The steering committee and the group
chairmen contributed additionally by selecting participants, orga-
nizing subject agenda, and refining reports produced during the
conferences. In particular, Drs. Richard, Hoar and Marshall Johnson
played major roles throughout the duration of the project. The final
step involved external reviews, and the contributors to this process
are also listed in the back of this report.
In addition to the input from the scientific community, efforts
of tne technical staff from Oak Ridge National Laboratory played a
large role in the success of the project. Joy Simmons and
Norma Callaham handled arrangements for hotel accommodations
and equipment rental. Debra Ballard, Evelyn Daniel, PatHartman,
Robert Ross, and John Smith provided word processing and logistical
support for the participants during the conferences. Members of the
Technical Publications Department, especially Pat Hartman
and Donna Stokes, were responsible for typing, and John Getsi,
for editing tne drafts into finished form.
The U.S. EPA's Offices of Research and Development and of
Pesticides and Toxic Substances and the Oak Ridge National
Laboratory gratefully acknowledge the efforts of all those involved
and thank tnem for assisting in this project.
Dr. Wayne Galbraith, U.S. EPA Co-Project Officer
Dr. Peter Voytek, U.S. EPA Co-Project Officer
Michael Ryon, ORNL Conference Coordinator
XI
-------
-------
CHAPTER 1
INTRODUCTION
The U.S. Environmental Protection Agency (EPA) has the
legislative mandate to consider regulatory alternatives for chemicals
that are causing or can cause a health hazard to man. Because the
reproductive system contains some of the more sensitive targets of
potentially hazardous agents whose impact on human populations
may be immediate, tpxicity to the reproductive system and the
conceptus is of emerging scientific and social interest. As a result of
this interest, the Offices of Health Research and of Health and
Environmental Assessment within the Office of Research and De-
velopment sponsored a conference to produce a technical document
on the current status of risk assessment methodologies for terato-
genic and other reproductive effects. The conference brought to-
gether scientists knowledgeable in reproductive biology and tera-
tology to discuss techniques and concepts pertinent to developing
risk assessment methodologies.
Conference participants were selected based on their expertise in
the various disciplines of reproductive biology, statistics, pharmaco-
kinetics, endocrinology, epidemiology, and sexual behavioral toxi-
cology. Draft copies of the report were sent to numerous scientists in
academia and the private sector for peer review, and their comments
were used by the members of the conference to modify the final
document.
The document is divided into three main subject areas: assess-
ment of toxicity to female reproduction, assessment of toxicity to
male reproduction, and assessment of toxicity to the conceptus.
There are three supplemental parts: pharmacokinetics and epide-
miologic considerations, which are common to all toxicological
assessments, and a special section on the behavioral aspects of sexual
development.
The specific areas addressed in this report are the potential
adverse effects on the female and male reproductive systems as well
as adverse effects on the developing conceptus. A broad range of
problems and effects are discussed, including infertility, early
resorption of the conceptus, and possible behavioral disorders
-------
produced by subtle changes in the biochemical environment of the
fetus.
The report also provides suggestions for improvement in standard
toxicological protocols for evaluation of reproductive risks, identifies
new concepts and procedures that can be immediately applicable,
and designates those that need further expansion and development
through research. Included is a discussion on the predictive ability of
the tests in estimating risk.
The information in this document will be of value not only to
scientists conducting experiments on the effects of chemical agents
on the reproductive system, but also to those that need to assess the
results from such studies. Thus many tests discussed herein may
currently be inappropriate, economically or technically, for regula-
tory use, but are included to provide necessary and useful back-
ground information for evaluating data.
In assessing human risks from exposure to potentially toxic
chemicals, many considerations should be addressed, such as severity
and reversibility/irreversibility of the effect, existence of threshold or
nonthreshold levels, dose-response relationships, sensitivity of the
toxicological response evaluated, and predictive ability of animal
studies to determine the risk to humans. Attempts have been made in
this document to address these considerations.
-------
CHAPTER 2
FEMALE REPRODUCTION
Risk assessment for toxicants that alter reproduction in females
involves two separate but equally critical tasks. These are assessment
of reproductive parameters in laboratory animals to identify com-
pounds that are prospective reproductive toxicants and continuous
epidemiologic surveillance of normal human reproductive character-
istics to identify their prevalence, trends, and geographical dif-
ferences and their potential modification by environmental events.
An approach to the evaluation of epidemiologic data is provided in
Chapter 5.
The problem of risk assessment has been approached by
proposing an animal screening system for qualitative and quantitative
analysis of reproductive toxicants. This system is coupled to an
integrated data base that serves as a mechanism for the analysis of
structure-function relationships of potential toxins. These testing
systems form a comprehensive screening scheme that should serve to
detect reproductive toxins and provide a foundation for risk
assessment. In addition, such a system will serve as a repository of
information into which continued input should expand our under-
standing of risk assessment and reproductive toxicology.
General Reproductive Toxicity Screen
We propose that the stepwise scheme shown in Fig. 1 be followed
in an attempt to identify compounds presently in the environment
for which there is epidemiologic evidence of adverse reproductive
effect and to identify new compounds that may be disseminated into
the environment. At the first level, a compound should be tested by
tne laboratory procedures described below. The standard acute,
subacute, and subchronic toxicological testing protocols do not
incorporate procedures for detection of reproductive effects, and
therefore the following screening procedures were specifically de-
signed for this purpose. If the result of any screening test is positive,
the compound must be evaluated by the quantitative risk assessment
procedures. If the screening tests are all negative, the compound
-------
ORNL-DWG 82-8409
Qualitative
toxicological
screen
Computerize
integrated
data base
Qualified
release
Quantitative
toxicologicai
screen
Release without
further
qualification
Risk assessment
decision
Figure 1 - Reproductive toxicity assessment prospective evaluation.
must then be compared by the computerized integrated data base for
structural relationships and/or similarities in the probable pharmaco-
kinetics witn other compounds known to affect the reproductive
system. An examination of the potential degradation products of the
compound, using a computer model analysis of its chemical
structure, might also prove useful. If the compound is found to have
a structural or functional similarity to known active agents, it must
undergo the quantitative risk assessment procedures. If no such
affinity is found (and the qualitative screen is entirely negative), the
compound can be released into the environment without further
testing.
-------
5
If a compound that enters the quantitative risk assessment
procedures is found to be without activity ("false positive" in screens
or in search for structural affinity), it can be given qualified approval.
Tnat is, its use must be restricted and accompanied by appropriately
directed epiaemiologic surveillance.
If a compound that enters the quantitative risk assessment
procedure is positive, the risk for humans should be estimated,
insofar as possible. This information must be weighed and the
decision made about whether the compound can be released at all
and if so, with what restrictions.
Qualitative Reproductive Toxicity Tests
Estrogen agonist—antagonist
Estrogens mediate, integrate, and modulate interaction of the
hypothalamic-hypophyseal-gonadal axis and, as such, are important
hormones in the control of reproduction. Exposure to exogenous
estrogens is known to have deleterious effects on reproductive
potential (1). For predicting the estrogenicity of environmental
chemicals, a series of simple screening tests are proposed. These
include (a) time of vaginal opening in the neonatal rat, (b) uterine
epithelial cellular hypertrophy, and (c) estrogen-receptor-binding
analysis. These tests have been chosen because of their sensitivity to
estrogenlike substances and the close correlations that exist between
these estrogenic responses and subsequent abnormal reproductive
capacity. For details of the test procedures, see the Appendix to this
chapter.
The tests cited above can be used to detect estrogen toxicants;
however, they could also be used to classify estrogen agonist-
antagonist. Generally, a decreased response in the tests would
indicate an antagonistic effect, whereas an increased response would
indicate an agonistic effect. Such a classification scheme, which
would require extension and expansion of the tests cited, could form
the basis for a structure-function data bank for the prediction of
estrogenic toxicity.
Androgen agonist—antagonist
Androgenic substances are known to cause infertility in female
animals, and their effects on the human fetus are well known.
Exposure to androgens during pregnancy causes masculinization of
female fetuses and various physiological and behavioral problems in
-------
the adult. Androgenic compounds can be assessed by their capacity
to stimulate weight increases in the ventral prostate and seminal
vesicles of the intact immature male rat or mouse. This assay is
acceptable and practical for predicting androgenic effects (agonistic
and/or antagonistic) in humans.
It is important to determine whether a potential toxicant
influences development or reproductive capabilities. For this pur-
pose, determining responses of newborn rats following exposure to
suitable doses of the potential toxicant provides a multifactorial
assay. Similarly, screening tests (e.g., testosterone blood levels or
accessory sex gland weight) for androgen antagonists are also
available. More sensitive tests for androgen agonists/antagonists are
described in the quantitative section of Female Reproduction and in
the Appendix to this chapter.
Nonsteroidal toxicant screening tests
The preceding tests will detect estrogenic and androgenic
toxicants. For testing of substances other than these two classes of
compounds, a multigenerational protocol is proposed. This protocol
is designed to evaluate (a) adult female conceptive ability with initial
exposure to the agent occurring near puberty, (b) the effect on
pregnancy, (c) potential transmission during lactation, and (d)
reproductive performance of the second generation. As a standard
approach to the testing of potential reproductive toxicants, these
tests will detect substances that interfere with reproduction at
various levels of biological organization. (For details see the
Appendix.)
Computerized integrated data base
The computerized integrated data base should include all known
structure-function relationships for reproductive toxicants. With such
a data base it would be possible to construct reproductive toxicant
profiles (activity profiles) that would predict the potential activity of
putative toxicants. Admittedly, such a scheme has shortcomings and
prediction will not be perfect. However, if at some future date
sufficient information were available in the data base, it could prove
to be most useful and time saving.
Quantitative Reproductive Toxicity Tests
The screening tests outlined in the previous section are designed
to identify compounds that may represent reproductive hazards.
-------
7
Once a compound is found to have either a reproductive effect in the
qualitative screen or a structural relationship with known active
agents in the computerized-integrated-data-base screen, a more
detailed quantitative evaluation is mandatory. This process will vary
substantially, depending on the type of effects seen in the qualitative
screen or the characteristics of the known toxicant to which it
appears similar. These tests for quantitative assessment, which are
presented in detail in the Appendix to this chapter, can be used to
determine site or locus of action of these xenobiotics. It is important
to recognize that some xenobiotics may act at more than one site
and by more than one mechanism. Ultimately, these specific assays
have the potential to determine risk of exposures. In this document
an attempt is made to provide an interface between female
reproductive biology and toxicology.
Risk Assessment
Assessment of risks to the female reproductive system from
environmental sources will have to involve a broad class of
potentially affected processes, organs, and structures obtained from
human exposure and relevant laboratory results. The reversibility of
effects needs to be considered carefully. The applicability of the
available information to potential environmental exposure will need
to be considered. The magnitude of human risk for reproductive
toxicity may be modulated by such diverse factors as distribution of
the compound in the environment, patterns of use or exposure,
persistence in the biosphere, concentration in the food chain, and
age-dependent changes in sensitivity.
Risk assessment will require a knowledge of pertinent factors
related to the reproductive process and of relationships of specialized
laboratory results to these factors. If a compound demonstrates a
reproductive effect in any mammalian species, this observation
indicates that some concern about actual human exposure to the
agent is justified. Positive results in a number of laboratory tests,
which by themselves may be only suggestive of harm, will be
important in evaluating potentially detrimental effects.
Substantial modifications in any of the subsystems given in
Table 1 are known to be serious and should be avoided. Future
testing may indicate relationships between these subsystems and
other laboratory testing. Risk assessment for female reproduction
requires the establishment of assays relevant to these reproductive
processes and the validation of these assays in identifying substances
actually toxic to human reproduction. The assays should be shown
-------
TABLE 1 Reproductive Processes Potentially
Susceptible to Reproductive Toxicants
Nonpregnant
Pregnant
Vulva/Vagina Viiilization
Adenosis
Cervix Structural abnormalities
Mucus production and/or
quality
Uterus Luminal fluid
Structural malformations
Dysfunctional bleeding
Dyssynergia
Deficient pseudodecidual
response
Fallopian Tube Gamete transport fluid
Ovary Decreased number of oocytes
Increased rate of follicular
atresia
Follicular: steroidogenesis
maturation
rupture
fluid quality
Oocyte maturation
Luteal function
Chronic anovulation
Breast Supernumerary mammary glands
Galactorrhea
Nongalactorrheic discharge
Gynecomastia
Placenta
Pituitary Hyperprolactinemia
Hypoprolactinemia
Altered synthesis and
release of trophic
hormones
Hypothalamus Altered synthesis and
release of neurotransmitters,
neuromodulators, and
neurohormones
Liver Metabolism
Binding protein synthesis
Adrenal Steroidogenesis
Behavior Sexual behavior
Reproductive Puberty
lifespan Menopause
Incompetence
Untimely parturition
Dysfunctional labor
Uterine blood flow
Gestational trophoblastic
disease
Deficient decidual response
Zygote transport
Ectopic pregnancy
Luteal function
Lactational transport of
toxicants
Lactation: composition
capability
Transplacental transport
of toxicants
Hydatidiform mole
Enzymatic activities
Metabolism
Binding protein synthesis
Steroidogenesis
Maternal behavior
-------
to be relevant in the species used, especially any assays adapted from
human assays, and should avoid complications such as those resulting
from diurnal variations in hormone level. The results from these
studies, may indicate other areas where regulation will be necessary.
It is generally felt that when dose-response relationships are
observed, lowering exposure will cause less harm when the com-
pound is xenobiotic, unless evidence to the contrary exists. Con-
sideration could be given to using safety factors to establish
acceptable exposure levels in situations where harm can reasonably
be expected and the exposure cannot reasonably be avoided.
Research Needed
Qualitative reproductive toxicity tests
The scheme for the detection of reproductive toxicity discussed
earlier and diagramed in Fig. 1 proposes further investigation in
several research areas. One such area is that of structure-function
relationships, which are not well understood at the present time. No
one would have predicted from the structure of kepone that it would
bind to the estrogen receptors and stimulate estrogenic responses.
Obviously, much needs to be learned about what constitutes an
estrogenic molecule. However, kepone would have been detected as
an estrogen by the above tests, and indeed, had more been known
about structure-function relationships, it might have been suspected
before any tests were performed.
The establishment of reproductive-toxicological profiles and
structure-function prediction models has just begun. Much basic
information is required before such a system can be realized.
Therefore, a strong recommendation is that basic research in
reproductive toxicity be supported, with a major emphasis on
establishing such models.
An important component of the qualitative reproductive toxicity
screen is the computerized integrated data base (see Fig. 1). With
such a data base, it should be possible to predict the potential
toxicity of putative toxicants. That such a predictive scheme has its
faults is well recognized; however, further efforts to realize the
potential of such a system should not be discouraged on account of
these. In theory, when sufficient information is available concerning
structure-function relationships of toxicants, such a prediction
scheme may decrease.the need for extensive animal testing. For this
reason it is recommended that further attempts to establish and
-------
10
validate such a data base be made and also that arrangements be
made to update continuously such a facility, with the ultimate intent
of perfecting predictive potential.
Quantitative reproductive toxicity tests
The recommendations concerning the qualitative tests also apply
to the proposed quantitative tests. The information gained from the
quantitative tests extend and interplay with the results obtained
from the qualitative tests. Therefore, it is recommended that
information gained from the quantitative tests be integrated with
that obtained from the qualitative tests to permit an even greater
understanding of structure-function relationships.
Specific recommendations
1. The relationship between cellular receptors for toxicants and
their mechanism of action should be explored further. Such
information can be fed directly into a structure-function data bank
such as the computerized integrated data base.
2. More work is obviously needed regarding masculinization of
the female, an important problem in reproductive toxicity evalua-
tions. Few data are available on dose responses of these effects, and
fewer data exist regarding inhibition (antagonism) of the alteration.
Further, extrapolation of these data to humans is not possible,
because subhuman primates and humans do not sustain substantial
defects of ovulation, whereas sexual behavior is altered. Information
presently available is insufficient for determining whether this
discrepancy is due to the fetal age at which treatment was
administered or to actual differences in sensitivities.
3. In vitro model systems are needed (in many areas) for the
assessment of reproductive toxicants. For example, model systems
for the secretion of gonadotropins by the pituitary cells can be used
to study toxicants that influence this process. Currently, almost
nothing is known about such model systems, and their value for
predicting toxicity is potentially great.
Another important in vitro model system in need of development
is that of inhibition of steroidogenesis. Although this system has
been well characterized for many inhibitors (see the Appendix to this
chapter), it has not been exploited for its potential as a test system
for toxicants. Continued work and support will be needed to develop
these model systems and to relate the information obtained to that
gathered from in vivo studies.
-------
4. Because little is known about the effects of neuroactive
substances such as dopamine and norephinephrine on the
hypothalamic-hypophysial complex and about the effects of toxi-
cants on this system, research support should be allocated to this
important field, both for the development of new methods and for
studies on the mechanism of action.
5. Support is recommended for development of models and
research on basic mechanisms in behavioral toxicology, an area in
which many unknowns exist regarding reproductive toxicity (see the
Appendix).
6. Much work is needed in oocyte toxicity, an obviously
important area of concern in which there are incompletely under-
stood age, strain, and species differences in sensitivity of ovotoxicity.
For example, preovulatpry and growing follicles are most sensitive to
toxicity in humans, whereas resting follicles are most sensitive in
mice. Similarly, significant differences in sensitivity to oocyte
destruction exist between mice and rats. However, evidence from
studies exploring the effects of antitumor agents on humans and
experimental animals suggests that a compound demonstrated to be
ovotoxic in rodents will also be ovotoxic in humans.
Extrapolation of animal data to humans
The primary goal of risk assessment for environmental agents is
directed toward adverse effect(s) (injury) in human individuals or
human populations. In most cases data are available only in animal
model systems; hence it is necessary to extrapolate these findings to
anticipated changes in humans. Although extrapolations may be
possible, it should be noted that our current understanding of the
relationships between hormone exposure and toxic outcomes is not
optimal. The following discussion is included to illustrate this point.
An increased rate of vascular disease in women taking oral
contraceptive pills has been reported by several investigators (see Kay
[2] for review). This has been generally attributed to the estrogenic
component of the pill and at first may seem to represent a source of
data concerning estrogen levels and toxic effects. However, as Kay
(2) points out, the progesterone content of the pill, not the estrogen
content, is correlated with increased incidence of vascular disease.
Progesterone has also been shown to decrease high-density lipo-
protein (HDL) cholesterol, an event associated with increased risk of
arteriosclerosis. Estrogens increase HDL cholesterol and therefore
would be expected to decrease the incidence of vascular disease.
Obviously, predicting risks based on estrogen levels in women taking
the pill requires further consideration.
-------
12
t
Tliis example points out the need for more research at all levels,
from biochemical to epidemiological, and emphasizes the require-
ment for more data before meaningful extrapolations can be made
for risk assessment in humans.
REFERENCES
1. McLachlan, J.: Estrogens in the environment. Elsevier/North Holland: New
York; 419 pp., 1980.
2. Kay, C. R.: The happiness pill? J. R. Coll. Gen. Pract. 30: 8-19, 1980.
-------
APPENDIX
DETAILS OF TEST PROTOCOLS AND
GLOSSARY OF TERMS FOR
FEMALE RISK ASSESSMENTS
I. DESCRIPTION AND DISCUSSION OF TESTS
USEFUL IN ASSESSING RISK TO THE
FEMALE REPRODUCTIVE SYSTEM
Qualitative Reproductive Toxicity Tests
Estrogen agonist—antagonist
Time of vaginal opening in the neonatal rat. Rats are injected on
days 1,3, and 5 of postnatal life, and the time of vaginal opening is
noted. Estrogen agonists such as diethylstilbestrol (DBS), clomi-
phene, and tamoxifen are known to cause early maturation of vaginal
development (1—3), and this test serves as a sensitive indication of
such activity. The general protocol for this test for estrogen agonist is
shown in Table 2.
Uterine epithelial cellular hypertrophy. Neonatal rats are treated
as described in Table 2, and the uteri are taken on day 7 for
histological examination. Epithelial cell growth is an excellent
indicator of estrogenicity and will detect compounds, such as
clomiphene, which exhibit differential cell stimulation (4). Kepone,
DES, dichlorodiphenyltrichloroethane (DDT), and zearlenone have
been shown to be either active in this test or very likely active
because of their known ability to stimulate uterine growth (4, 5).
The above tests, requiring a minimal number of animals, are
simple and reliable. These tests are used routinely in many
laboratories and are quite sensitive to estrogenic compounds (/zg/kg).
Estrogen-receptor-binding analysis. Uteri obtained from 7-day-
old rats which have been treated as described above are examined for
nuclear binding of the estrogen receptor by the nuclear exchange
assay (4). In the same tissues the quantity of cytoplasmic estrogen
receptor can also be determined. This test gives a measure of the
13
-------
14
TABLE 2 Estrogen Agonist Screen
Age
(days)
Treatment of
female rats
0 Birth
4 Beginning of daily dosing for 4 days
7 Sacrifice
Assays:
1 — uterine weight
2 — endometrial histology
3 — estrogen receptor assays
4 - vaginal opening0
"Vaginal opening may occur by day 7; however,
a longer time interval after birth may be required
(up to 20 days).
ability of a toxicant to bind to estrogen receptor in vivo and to cause
nuclear accumulation of the receptor-ligand complex. Classical
estrogens such as estradiol and DBS are known to perform this
function, which is presumed to be an obligatory step in the
mechanism of action of estrogens. Kepone, DDT, and zearlenone
bind to the estrogen receptors, cause nuclear accumulation, and
stimulate uterine growth (3, 5 6). Therefore, these compounds are
likely to be active in the other tests for estrogenicity and will make
excellent reference compounds for testing the model.
Aiidrogen agonist—antagonist
These tests are standard and require no further explanation.
Nonsteroidal toxicant screening tests
This protocol is designed to evaluate: (a) ability of the adult
female to conceive with the initial exposure to the agent near
puberty (P-generation, Table 3); (b) the effect on pregnancy (live
birth index of Fj, F\, F2, F2 generations); (c) potential transmission
during lactation (survival index of Fj and F2 generations); and (d)
reproductive performance of the second generation (live birth index
of Fj and F2 generations). Part of the P, generation (PJ) is mated
again at the time of postpartum estrus, because at that time mating
behavior, ovulation, implantation, and fetal resorption are more
sensitive to environmental disruption than they are during mating at
a cycling estrus.
-------
15
o
W
c
s
o
0>
1 „
(0 g
i'l
w 3
o £j
s i
O 3
•a W
2 £
-2 'S
£3 -2
l|
•2 a.
.0 C
«w o
1
1
uT
tf
»> •« ri
.S E — fc
Iffe l^
**« •*"" -"K C T3
O *-> '~* o C
13 H M " °
"^ ^~* sn w So
HI I i
^ C X> G 43
III II
tiT uT ciT oT uT
"O
ZJ
.s
'c ^
^o rt
I 3 1 1
QJ ^ •*-• "^
op ^3 M a, tL<
c .E S 2 » S
i^ w aj ca O rt
1 1 1 1 (JI 1
B. CL, CL, CL, CL, ft.
Of*^ \om»oo\O<^
^^T T7T7TT
.i
| oil
ll
-------
16
The age at which the animals are mated is a major procedural
factor that can Wluence the fertility test results in the F1 and F2
generations. Toxicants, particularly those that possess steroidal
activity, will diminish the success of pregnancy and number of
offspring of older females but not those of younger females. Thus,
while testing at earlier ages is more economical, it might yield false
negative results. Therefore, a portion of the Fj generation should be
examined for ovarian cyclicity and fertility at approximately six
months of age (live birth index of F^ generation).
Experimental, vehicle control, and positive control (use of a
known toxicant to verify the system) groups should be utilized with
at least 20 animals per group. Selecting the agent to be used as a
positive control will be arbitrary, and species or strain differences
may complicate the choice. Despite these drawbacks, inclusion of a
positive control that most appropriately parallels the test compound
would seem mandatory for validation of the test system. The entire
protocol need not be completed if adverse effects are demonstrated
early in the protocol (i.e., live birth index of the Fj generation).
The maximum tolerated dose should be used. Other dose levels
may be included if dose response information is needed. Route of
administration should be in food or water to avoid handling pregnant
and lactating females, which may result in stress independent of that
potentially caused by the agent being tested. This may complicate
quantification of the ingested dose but ensures continuous dosing of
the Fj generation during weaning. Other routes of exposure (e.g.,
gavage or parenteral administration) may be used if the test protocol
can be modified to avoid any interfering stress. Dosing begins at six
weeks of age of the Pt generation and continues until the end of the
protocol. Body weights should be recorded weekly for all animals in
the P! and F: generations as well as pup weights in the Fj, FJ, F2,
and F2 offspring.
P! females are mated with untreated males of proven fertility at
ten weeks of age in a one-to-one sex ratio. Successful mating is
determined by finding a copulation plug and presence of sperm in
the vaginal smear. These same females are then mated again at the
time of postpartum estrus, 8-10 hours after giving birth. Twenty
females from the Fl generation (the offspring resulting from the first
mating), are randomly selected and mated with untreated males of
proven fertility at ten weeks of age. The offspring of postpartum
mating (FJ) need only be counted and weighed at birth.
The selection of the species of the test animal to be used in the
toxicant screening procedures will be determined by several con-
siderations including cost, time, and ability to assess related human
-------
17
reproductive processes. The rhesus monkey or other subhuman
primates, being a more comparable reproductive species, would be
the animals of choice, but their cost as well as other considerations
would prohibit their use in screening procedures. Laboratory rodents
are economically feasible, but the relevance of the outcome of the
screening to the human could be questioned. On the basis of current
information, different species and strains will have to be selected for
evaluating different components of the human reproductive system.
Each choice would carry with it a risk of obtaining false positive and
false negative data with regard to the relevance to the human female.
For example, on the basis of contemporary results, the rat would be
less satisfactory than the guinea pig for assessing the effects of
potential toxicants on the development or the integrity of cyclic
gonadotropic function.
Several neural and physiological interventions that curtail estrous
cycling in the rat do not occur in the rhesus monkey and guinea pig.
In addition at least some perinatal steroid manipulations that render
the rat permanently anovulatory apparently do not interfere with
menstrual cycles in the rhesus monkey. Thus, it is likely that many
substances found to disrupt spontaneous ovulation in the rat will not
do so in the human, and false positive assessments may result.
A false negative may occur if the rat is the only species used to
assess the reproductive consequences of a compound. For example,
the ovarian cycle of the rat does not have a spontaneous luteal phase
as does the human cycle. Therefore, compounds that might interfere
with the function of the corpora lutea cannot be detected in the rat.
Under these circumstances another species with a comparable
reproductive process, such as the guinea pig, should be considered for
addition to the screening procedure. .
Indexes should be calculated for mating, fecundity, female
fertility, and parturition as noted below.
mating index = numbe? of copulations (one counted/estrous cycle)
number of estrous cycles required
x 100
fecundity index =
numer
number of copulations
fertility index = numb^r of fi^f conceiving
number of females exposed
incidence of parturition = num^er °£ Parturitions
number of pregnancies
-------
18
Numbers of viable, stillborn, and cannibalized progeny are
recorded for each litter, the survivors on days 1, 4, 12, and 21
postpartum noted, and litters reduced to ten pups on the fourth day
of lactation for standardization. Gestational length and sex ratio are
also monitored.
,.,,., number of viable pups born 1fin
live birth index = total number of pups born x 1 °°
. , number of pups viable at location day 1 or 4
1- or 4-day survival = number of pups born
index
_, , . , _ number of pups viable at lactation day 12 or 21
12- or 21-day survival - number of pups retained at lactation day 4
index
Quantitative Reproductive Toxicity Tests
Estrogen agonist—antagonist
The screening test listed previously under qualitative assessment
can be used to establish dose-response relationships between estro-
gens and suspected estrogenic toxicants. The following discussion
represents an expansion of the qualitative tests.
Neonatal exposure to various dose levels of estrogenic toxicants.
These assays will result in a dose-response relationship for time of
vaginal opening, ovarian degeneration and oocyte loss, and stimula-
tion of epithelial cell height in the uterus. These end points are easy
to assess, are reproducible, and are quite sensitive (jug quantities of
DBS, Kepone, and clomiphene are easily detected) (1-3). This is not
to say that these tests have been utilized to examine a large class of
compounds; however, one of the recommendations is that such
compounds be studied in detail. At the present time all known
estrogens are active in these assays, and hence we can expect that
they will be good predictors of estrogenic potency. Likewise, such
assays should identify compounds that may interfere with reproduc-
tive processes. It may be possible to extrapolate these data on
relative potency to known effects of various doses of estrogens in the
human, since it is well established that estrogenic responses in
rodents and humans show many similarities (7—9). To this end
compounds such as ethynylestradiol, DBS, and estradiol should be
used as standards.
Estrogen receptor analysis in vivo and in vitro. An extremely
sensitive (picogram-nanogram range) and reproducible method for
-------
19
assessing relative estrogenicity involves the use of toxicants to
compete with labeled estradiol in binding to uterine cytoplasmic
estrogen receptors (3-6). This test involves the addition of various
concentrations of the toxicant to uterine cytosol fractions in the
presence of labeled estradiol. If the toxicant is estrogenic, it will
compete with estradiol for binding to receptor sites, and a classical
competitive inhibition curve can be obtained. From this curve a
relative binding affinity (RBA) can be calculated that reflects the
agonistic or antagonistic activity of the toxicant. Such estimates of
potential estrogenicity may be used to extrapolate estrogenicity in
humans and be of importance in approximating the relative risks in
humans. Although this test is simple and requires little expense in
terms of number of animals, etc., not all laboratories routinely
perform such analyses. However, it is becoming more and more
common and may be standard procedure in the future.
The major qualifier to such cytosol receptor assays is that certain
estrogenic compounds, such as clomiphene and nafoxidine, exhibit a
very low RBA-and yet are more estrogenic than predicted (10). In
part this is due to the slow clearance of such compounds, which
provides a longer exposure time and increases the receptor occu-
pancy in vivo when compared to that of more rapidly cleared
estrogens. To detect such long-acting estrogens, estrogen receptors
assays can be done in vivo. Mentioned in the section titled
Qualitative Reproductive Toxicity Screen, these assays involve
injecting various dose levels of the compound in immature rats and
measuring the nuclear accumulation and cytoplasmic depletion of
estrogen receptors. Reliable, easy to perform, and sensitive, this test
requires relatively few animals. It has the disadvantage of not being a
standard assay in all laboratories.
Such receptor assays can be valuable in the estimation of
estrogenic potency in humans; however, the chief value of the
receptor assay probably lies in its ability to detect estrogen agonist or
antagonist and has the potential of elucidating primary steps in the
mechanism of action of such compounds. Such insights into
mechanisms may make future predictions of estrogenic toxicity a
relatively simple task.
Androgen agonist—antagonist
Qualitative screening tests for androgen activity include the
ventral prostate gland hypertrophy produced by administration of
compounds to immature (28-day) male rats.
-------
20
Additional models are necessary to evaluate androgenic effects in
the following circumstances: (a) inhibition of adult female reproduc-
tive function (e.g., ovulation, behavior), and (b) masculinization of
female phenotype (fetal differentiation, prepubertal development,
and adulthood phenotypic transformations).
Inhibition of adult female reproductive function. The common
clinical response to hyperandrogenic stimuli, is anovulation. In-
creasing duration of exposure or potency of the agent leads to
oligomenorrhea and secondary amenorrhea. Subtle intensifications of
libido are experienced by some women, particularly with more
potent agents.
Laboratory testing of adult female rats requires daily evaluation
of vaginal smears for no fewer than four cycles to detect interruption
of the estrous cycle. A daily 1-mg dose of testosterone propionate
produces diestrus within two cycles. Appropriate dose-response
studies are indicated.
Masculinization of the female phenotype: fetal. After 16 days
of gestation, transplacental transfer of potent androgen agonists
results in a variety of imprinting and masculinizing responses that are
based upon "critical periods" of organ system differentiation.
Permanent alterations in the neuroendocrine regulation of the
estrous cycle and male-type mating behavior are "imprinted" at
lower doses of androgen than are required for disturbing reproduc-
tive tract (vaginal opening) and hepatic monooxygenase (steroid
hydroxylase or dehydrogenase) activities. A 5-mg dose of testos-
terone propionate administered to the pregnant dam daily from day
16 to day 20 of gestation produces the masculinization response in
female progeny and does not significantly disturb male differentia-
tion. Treatment of neonatal female rats (day 1-10) with a single
1-mg dose of testosterone propionate masculinizes the hypotha-
lamic-pituitary-ovarian axis (persistent estrus) and sexual behavior
(male-type with great reliability).
Masculinization of the female phenotype: postnatal ani-
mals. Masculinization of the female phenotype and suppression of
female sexual behavior and of the pubertal events is not induced
permanently by treatments initiated after the postnatal period (days
1—10). Such masculinization effects produced in females tend to
regress, and although vulvar changes may persist, estrous cyclicity
resumes. Although these effects are clear-cut in rodents, dose
extrapolation to humans is not possible.
Adulthood phenotypic transformations. Masculinization of
vulva, mating behavior, and hepatic monooxygenases in adult animals
-------
21
are much less sensitive indicators of toxicity than the same responses
in immature animals. Inhibition of ovulation in adult females is also a
more sensitive indicator of toxicity than the above three parameters.
Therefore, additional tests to assess phenotypic transformation in
adult females are not necessary.
Hypothalamic—Pituitary Function Tests
Assay of agents that stimulate the release of gonadotropins
from cells of the anterior pituitary gland
A toxicant may adversely affect reproduction by altering the rate
of secretion of one or more hormones that are synthesized and
released by the hypothalamus and anterior pituitary gland. Of the
hormones that are secreted by the anterior pituitary gland, the
gonadotropins (luteinizing hormone [LH] and follicle-stimulating
hormone [FSH]) and prolactin are most closely associated with
reproduction. The gonadotropins are important, because these
protein hormones control ovarian function, including steroid hor-
mone secretion, follicular development, and' ovulation. Hence, if
gonadotropin secretion is suppressed, ovarian function is suppressed.
A toxicant could suppress the secretion of gonadotropins by acting
directly on the pituitary gland or by suppressing the secretion of
gonadotropin-releasing hormone (GnRH) by hypothalamic neurons.
Alternatively, a toxicant could stimulate the secretion of
prolactin, and as a consequence of the hyperprolactinemia, gonado-
tropin secretion becomes suppressed. Prolactin secretion can be
stimulated by substances that have estrogenic activity, substances
that act as dopamine antagonists, substances that inhibit dopamine
secretion by hypothalamic dopaminergic neurons, or substances that
cause hyperplasia of prolactin-secreting cells. Some of these actions
of toxicants can be assessed (e.g., by quantifying gonadotropin and
prolactin secretion), whereas others cannot be evaluated in a
quantitative sense (e.g., GnRH secretion). A few ways of assessing
quantitatively the actions of a toxicant that may have significant
effects on reproduction are listed below.
In vivo model. Since agents that stimulate the release of one
gonadotropin (e.g., LH) usually affect the release of the other (viz.,
FSH), it is probably only necessary to measure the release of one
(e.g., LH). For such studies, the estrogen-progesterone-primed female
rat can be used. The Gn-releasing standard should be synthetic GnRH
against which the test substance can be compared. The responsive
parameter, LH in serum or plasma, can be measured by a
-------
22
standardized radioimmunoassay. After the ED50 of GnRH and the
ED go of the test substance have been ascertained, a standard
bioassay can be performed. A three-dose assay, where each dose is
replicated three to four times, may suffice. Thus, such an assay will
require 18 to 24 assay animals. More animals can be used if high
precision is desired. The concentration of LH in plasma or serum can
be evaluated 30 to 60 minutes after the administration of the test
substance and of the standard. Assuming suitable ranges of concen-
tration and parallel slopes from the two assays and a suitable
statistical analysis, such as that described by Bliss (11) for bioassays,
qualitative characteristics of the assay as well as the relative potency
of the unknown substance can be evaluated. If one knows the
potency of GnRH in the estrogenized woman with reference to the
release of LH, it is then possible to calculate the Gn-releasing activity
of the test substance and express the potency in terms of GnRH.
In vitro model. The Gn-releasing properties of an unknown
substance can also be measured using anterior pituitary cells
maintained in monolayer culture. In this case the pituitary cells
could be obtained from the rat or a suitable primate. The cells could
be dispersed and established in culture. After three to five days, the
test substance and GnRH can be assayed for their Gn-releasing
activities, using a bioassay paradigm similar to that outlined above.
Assays of agents that inhibit the release of gonadotropins from
cells of the anterior pituitary gland
The details of an in vivo model assay of a substance that inhibits
the release of gonadotropins could be done as follows. A female rat
castrated 4 to 6 weeks before testing could serve as the assay animal.
(In such an animal, the concentration of LH is many times that of
intact animals.) For a reference standard, 17j3-estradiol could be used
to suppress the release and hence concentration of LH in serum of
the test animal. After the ED50 for estradiol and the EDSO for the
test substance have been established, a three-dose bioassay could be
conducted. After an evaluation of the parameters of the assay, it may
be possible to calculate the relative potency of the test substance and
express its potency in terms of 17/3-estradiol.
After the potency of a test substance relative to 17|3-estradiol has
been established, one can then calculate the relative potency of the
test substance in the human by comparing the .LH-lowering effect of
170-estradiol in castrated or postmenopausal women.
-------
23
Assay of an agent that inhibits the release of prolactin from
cells of the anterior pituitary gland
In vivo model. Although several experimental animal models
could be used, the young, mature female rat would be adequate. The
only pretreatment required would be a short period (3-4 days)
during which the animal was handled to minimize the release of
prolactin because of fright. The reference standard for the release of
prolactin could be haloperidol. Prolactin release could be evaluated
by measuring prolactin in the plasma or serum of the test animal.
Thirty to sixty minutes after the administration of the test substance
or haloperidol, serum or plasma prolactin could be measured by
radioimmunoassay. After the dose-response curve (i.e., ED50) is
established, a bioassay could be conducted and the relative potency
of the test substance calculated.
If a dose-response curve for haloperidol in women is established,
one could approximate the potency of the test substance relative to
haloperidol. Of course, other reference standards could be used in
this prolactin release assay in vivo. These include thyrotropin-
releasing factor and vasoactive intestinal peptide.
In vitro model. The ability of test substance to simulate the
release of prolactin from pituitary cells could be conducted in vitro
using pituitary cells maintained in monolayer culture. The donor
could be the rat as well as a primate. After a few (3—5) days in
culture, a suitable bioassay could be performed, and the potency of
the test substance relative to a standard could be evaluated.
Assay of an agent that inhibits the release of prolactin from
pituitary cells
In vivo model. An estrogen-primed female rat could be used in
this assay. In such an animal the plasma concentration of prolactin is
very high. For a reference standard, bromoergocriptine could be
used. After dose-response curves for bromoergocriptine and for the
test substance had been established, a bioassay for prolactin release
inhibition could be performed, where the serum or plasma prolactin
concentration is the responsive variable. After a suitable statistical
analysis, one could calculate the relative potency of the test
substance. As discussed above, if the dose-response relationship for
bromoergocriptine in the woman (perhaps an estrogenized woman)
were known, the approximate potency of the test substance relative
to bromoergocriptine could be calculated.
In vitro model. Anterior pituitary tissue from estrogenized
female rats can be used under in vitro conditions to test a substance
-------
24
for inhibition of prolactin release. Anterior pituitary tissue can be
incubated in the presence of various concentrations of bromo-
ergocriptine or of the test substance to establish a dose-response
curve. Then, using this system, a bioassay can be conducted where
the concentration of prolactin in the culture medium is the
responsive variable. Pituitary cells maintained in monolayer culture
could also serve as a suitable in vitro assay system.
Assay of the activity of an agent that alters the secretion of
dopamine by hypothalamic neurons
There is no method for the quantification of the secretion by
dopaminergic neurons in the human. Although there is a method for
the measurement of the secretion of dopamine into hypophysial
portal blood, the procedure is tedious and requires the aid of highly
skilled people. Hence, this procedure is not practical as a routine
procedure. Therefore, one is reduced to making turnover measure-
ments, but such measurements are also susceptible to large error and
require many animals. Thus, it is reasonable to conclude that as a
routine matter, the rate of secretion of dopamine by neurons of the
brain can not be done for toxicants. This is not to infer that this is
not an important aspect of brain function. Indeed, it is already
known that the secretory activity by dopaminergic neurons is
quickly, markedly, and sometimes permanently affected by a variety
of toxicants. Since dopaminergic neurons constitute an important
subset of the neurons of the brain, we encourage research on this
important topic.
Assay of the activity of an agent that alters the secretion of
norepinephrine by hypothalamic neurons
Comments made about the secretion of dopaminergic neurons
are equally applicable to neurons that secrete such biogenic agents as
norepinephrine and serotonin. The available techniques for the
quantitative study of neurons secreting these agents are not
sufficiently advanced to enable their use in routine assays.
Assay of the activity of an agent that alters the secretion of GnRH
There is no suitable assay for such an agent at this time.
-------
25
Assay of the activity of an agent that alters the secretion of
hypothalamic opioid peptides
It is now clear that morphine and opiatelike peptides affect the
secretion of dopamine by hypothalamic neurons and of LH by the
pituitary gland. Thus, it is easy to infer an important role for the
naturally occurring opiatelike peptides as well as morphine in
reproduction. However, this field is too new to address in a
quantitative manner or to include in a screening system. Yet it can be
anticipated that at some time in the reasonable future, this
shortcoming in our technical capabilities will be surmounted and
these problems addressed in quantitative terms.
Blood flow of the hypothalamic-hypophysial system
Perhaps no structure in the mammal has a more complicated
vasculature than the hypothalamic-hypophysial complex, consisting
as it does of one component that has a high rate of perfusion and
another that is avascular. Moreover, the coexistence in the pituitary
stalk of portal vessels carrying blood to the anterior lobe of the
pituitary from the hypothalamus and of a subependymal plexus in
which pituitary hormones can pass retrograde in the stalk to the
hypothalamus attests to the importance of blood flow in this area.
The measurement of blood flow to the neurohypophysis (i.e.,
medium eminence and pars nervosa) can be measured accurately
using radiolabeled microspheres. Blood flow in the anterior lobe of
the pituitary can be measured using the hydrogen electrode. Thus, an
area deserving of attention for effects of environmental toxicants is
the hypothalamic-hypophysial complex.
Sexual behavior tests
Introduction. Alterations of mating behavior in the female rat
can be used as an indicator of hypothalamic function and/or
impairment of function. A voluminous literature indicates that
hypothalamic neurons serve as target cells for the ovarian hormones,
especially the estrogens, and that destruction of specific regions of
the hypothalamus leads to abolition or disorganization of female
sexual behavior (12). The studies carried out on the disturbance of
sexual behavior associated with hypothalamic damage establish three
points of significance to the use of the proposed sexual behavior
tests: hypothalamic damage can disrupt sexual behavior without
altering the neural systems that mediate pituitary-ovarian function,
the disruption of sexual behavior following hypothalamic damage
-------
26
cannot be reversed with endocrine therapy, and the sexual dis-
ruptions show extensive phylogenetic continuity.
For these reasons, the sexual behavior system may be capable of
detecting chemically induced abnormalities in hypothalamic function
that cannot be detected by the other testing systems proposed. Ly
substituting the putative toxicant for estrogen or progesterone in a
standardized behavioral assay, it is possible to assess its estrogenic or
progestogenic activity in hypothalamic regions other than those that
modulate pituitary release of gonadotropins. Neural systems modu-
lating female sexual behavior are by no means limited to hypotha-.
lamic structures. The tests proposed here, however, are oriented
toward behavioral end points generally accepted as involving the
hypothalamus. A more detailed discussion of other aspects of sexual
function and the potential toxicant-induced disruption of other
neural systems is available in Chapter 5.
Behavioral assay methods. Female rat sexual behavior has
several components that vary, in a dosage-dependent manner with
estrogens and progestins. To determine whether the toxicant
possesses estrogenic or progestagenic action, the experimenter would
vary the dosage of the toxicant, substituting it for either estradiol
benzoate (EB) or progesterone in the standard protocol of a behavior
assay. The lordosis response, which includes arching of the back (13),
and the number or latency of approaches that the female makes to
the male (14) can be readily quantified. The testing arena should
have an area of at least four square feet and contain a simple barrier
or compartment. The females used in the tests should be ovariec-
tomized and administered estrogen and progesterone (or the sub-
stituted toxicant) at times to produce mating during the dark phase
of illumination. The estrogen EB (or its substitute) is administered
44 to 46 hours before the administration of progesterone (or its
substitute), and the mating behavior is observed approximately four
hours after administering progesterone.
A standard dose-response curve for EB would be obtained by
holding the amount of progesterone constant at approximately 0.5
mg and varying the dosage of EB from 0.1 to 100 jitg. When possible,
the vehicle for delivering these hormones should be the same as that
to be used for the toxicant and appropriate standard curves
established. The progesterone standard dosage-response curve would
be obtained by maintaining EB at a constant level (5-10 Mg) and
varying the amount of progesterone from 0.1 to 10 mg. No fewer
than 10 subjects can be used to establish the behavioral response
value at each dosage. Repeated mating of the same female at each
-------
27
dosage may be more sensitive than using different females,at each
dosage, but each mating should be separated by approximately 5 to 7
days.
A reliable method to determine the dose-response curve for
antiestrogenic activity would be to use guinea pigs as subjects and to
measure their lordosis response to the touch of the experimenter;
running the index finger along the back, starting from its most caudal
point. The procedure would consist of administering constant
dosages of EB (5-10 pig) and, 46 hours after the initial dose of EB, a
0.5-mg dose of progesterone. When administered simultaneously with
estrogen, progesterone possesses antiestrogenic properties in the
guinea pig and can therefore be used as a standard to assess the
antiestrogenic activity of the toxicant. Accordingly, varying dosages
of the compound (0.1—10 mg) would be administered along with the
EB.
The antiestrogenicity of a putative toxicant can also be evaluated
in the female rat's behavior system. However, because this behavior
system is relatively insensitive to the inhibitory actions or anti-
estrogenic actions of progesterone, an alternative antiestrogen, such
as MER-25, is recommended for a comparative standard.
The toxicant can also be administered in addition to the standard
doses of EB or progesterone. This can be done at the same time that
estrogen and progesterone are administered to determine whether it
potentiates or antagonizes the action of each of these steroids. In
addition, the toxicant can be administered prior to the determination
of a standard dose-response curve, and the estrogen dose-response
curves can be compared with those obtained from control animals. In
this way, toxicant-produced modifications of the brain areas medi-
ating sexual behavior can be detected using a test based on a standard
estrogen dose-response curve.
Relevance to humans. A positive result from these behavioral
screening procedures could reflect disruption of neural, most likely
hypothalamic, function and could indicate the potential for inter-
ference with human hypothalamic function. However, the manifesta-
tions of this potential hypothalamic disruption in humans will most
likely be different from those in rodents.
Ovarian Toxicity
Oocyte and follicle toxicity
The ovary is responsible for two roles in reproduction: nurture
and release of gametes and hormone production. Clinical and
-------
28
experimental data demonstrate that a variety of xenobiotic com-
pounds can alter both aspects of ovarian function (15, 16). Multiple
studies have demonstrated that one of cigarette smoking's adverse
effects on the human ovary is an earlier dose-related age of
menopause. The assays described here are designed to assess the
effect of xenobiotics on the first aspect of ovarian function, gamete
nurture. Tests for xenobiotic effects on oogenesis can be determined
by including prenatal as well as postnatal treatments. Xenobiotic
destruction of oocytes is of great significance because the effect is
irreversible: there is no mechanism for repopulation of oocytes in the
ovary.
Evidence suggests that inbred mouse strains represent the most
sensitive test strains for oocyte and follicle toxicity assays (17).
Additional data in other species and with other xenobiotics is needed
to clarify this relationship. Inbred mouse strains also offer the
advantage that they provide the logical framework for exploration of
the mechanism of action as well as providing a reproducible assay
system (18-20).
After treatment with the compound of interest, mice are
sacrificed at varying time intervals and their ovaries removed, fixed,
serially sectioned, and stained. Oocytes and follicles are quantitated
using a microscope, and effect of treatment on oocyte or follicle
number is determined. Follicles and oocytes are classified by the
method of Zuckerman (21). This assay, although cumbersome, is
easily learned and conducted by laboratory technicians.
Evidence suggests that this assay is a much more sensitive
indicator of oocyte or follicle damage, than alterations in fertility.
Unpublished investigations at the Pregnancy Research Branch of the
National Institute of Child Health and Human Development, as well
as other published data, suggest that as many as 90% of all oocytes
have to be destroyed before short-term alterations in fertility of the
female can be observed.
The full range of specificity of the assay has yet to be
determined. The assay appears to be quite sensitive and dose
dependent (see Table 4 for available ED5 0 's).
Inhibition of steroidogenesis
In developing a model system to estimate the quantitative risk of
a toxicant with regard to inhibition of ovarian steroidogenesis,
multiple physiological and technical aspects must be considered.
These include (a) the cell-type specific sex steroids to be measured,
-------
29
Q £P
W g
CD
Compound
A A A A
— i
g
^
Cq ^
ti o
CQ +•>
c c
0) o
8 5
o o o o .ti !3 -tl
•*-< •»-! -T-l «1H S"* ? _Q
OOOOOOOOOOOOo45
CO
s
c
03
CD
CO
s
CD r-<
c §
(U )H
CD
CD
-5 S9 -5 « -S
'
1
s
CO
1
£
I
V3
CD
-------
30
(b) the cooperative compartmental steroid biosynthesis charac-
terizing the follicular phase, (c) the cycle-related variations in sex
steroid production, (d) the key regulatory steps in steroid bio-
synthesis (i.e., the availability of substrate and the roles of luteinizing
hormone [LH] and follicle-stimulating hormone [FSH]), (e) the
available methodology, and (f) the ability to extrapolate the data
between species. The focus here is placed on a model in vitro rather
than in vivo because of considerations regarding the specificity of the
toxic effect, the lack of interference by nervous or humoral factors
present in vivo, the greater likelihood of intracellular interaction with
the toxicant, the applicability of the test system, and the ability to
extrapolate the data to human or at least primate ovarian cell types.
Estrogen, primarily 17/3-estradiol progesterone, 17a-OH proges-
terone, androstenedione, and testosterone, are the predominant
steroids produced by the human ovary during the reproductive years.
Estrogen characterizes the follicular phase, with the corpus luteum
producing both estrogen and progesterone and a drop in both
steroids occurring at the time of menses in a nonconceptive cycle.
Androgens are secreted throughout a nonconceptive cycle, with a
slight rise at midcycle. Controversy still exists concerning the cell(s)
of origin of follicular estrogen; both direct thecal cell secretion and
granulosa cell aromatization of thecal androgen are supported in the
literature (23). Because granulosa cells lack the 17,20 desmolase
enzyme, the thecal and interstitial compartments are felt to be the
source of Ct 9 androgens. After ovulation the granulosa and thecal
compartments both form the corpus luteum and produce proges-
terone and estrogen. Any model system using ovarian cell types in
vitro must consider these differences as well as the overall cyclic
steroid secretory pattern characteristic to the species utilized.
Regulatory steps in gonadal steroid secretion include (a) sub-
strate (cholesterol) availability (i.e., the low-density lipoprotein
fraction of plasma); (b) luteinizing-hormone (LH) induction of the
20,22-hydroxylase-desmolase steps converting cholesterol to preg-
nenolone, and (c) follicle-stimulating-hormone (FSH) induction of
granulosa cell aromatase activity converting thecal androgens to
estrogens. Since thecal steroidogenesis has not been demonstrated to
depend on FSH-induced aromatization, LH stimulates thecal andro-
gen production, and low levels of LH are required in vivo for
adequate luteal function, some of these regulatory steps may be
compartment specific.
A toxicant may not demonstrate inhibition of steroidogenesis in
vitro and yet be active in vivo, if it affects selectively gonadotropin-
mediated events in vivo or only progesterone synthesis stimulated by
-------
31
human .chorionic gonadotropin (HCG) (24, 25), and hence be
detectable only in vivo in a known conceptive cycle. Similarly, agents
acting through prostaglandins known to induce luteal regression in
vivo in some species (26) may be active only in vivo, because the
agents may act not directly on the steroid-secreting cell but rather
indirectly by selective ovarian veno-constrictive action. Despite these
possibilities, most known inhibitors of steroidogenesis act by
affecting specific enzymes in the steroid pathways (Table 5).
TABLE 5 Agents That Inhibit Steioidogenesis
Steroidogenic Step
20a hydroxylase
Sidechain cleavage
Dehydrogenase, 3(3-hydroxy-
A* -steroid
Aromatase
11/3-hydroxylase
21-hydroxylase
17a-hydroxylase
17,201yase
' Inhibitor
Amino-glutethimide phosphate
3-methoxybenzidine
Cyanoketone
Estrogens
Azastene
Danazol
4-acetoxy-androstene-3,17-dione
4-hydroxy-androstene-3,17-dione
l,4,6-androstatriene-3,17-dione
Danazol
Metyrapone
SKF-12185
Danazol
Danazol
SU-9055
.SU-8000
Danazol
Adequate methodology is currently available for (a) isolation of
ovarian cell types (27, 28), (b) tissue or organ culture, and (c) direct
radioimmunoassay of media for individual steroids without chroma-
tography steps. If HCG stimulation of steroidogenesis is required to
demonstrate an effect, serum-free media may be required, as there is
some evidence that blocking factors for gonadotropins are present in
serum (29); but in short-term cultures (<24 hours), the lack of serum
factors should not present a problem for cell viability. Plating
efficiency can be determined by supravital staining, and cell counts
or determinations of DNA or protein can be used to normalize data.
Organ cultures are more difficult to normalize because of more
heterogeneous cell populations, less well-defined culture conditions,
and more difficult assessments of cell viability, but tissue wet weights
can be used. Enzymatic dispersion techniques are available (25), but
-------
32
they add considerable time to the procedure and do not solve the
problem of cell heterogenicity. Furthermore, if gonadotropin stimu-
lation is required, highly purified enzyme preparations (i.e., col-
lagenase) are necessary to avoid protese contamination and altera-
tions in membrane-bound protein receptors (30). Hence, because of
ease of culture, purity of cell type, and active basal steroidogenesis,
isolated granulosa cell cultures, with or without added Ci 9 androgen
substrate, represent attractive models for evaluation of a potential
toxicant's effect on steroidogenesis (Table 6).
Cell-cell interactions may control the pattern of ovarian steroido-
genesis as evidenced by the so-called "spontaneous luteinization"
that granulosa cells undergo when placed in tissue culture inde-
pendent of when they are harvested in the follicular phase (31). The
removal of the cells from their approximation to the thecal layer,
contact with follicular fluid, or disruption of intimate cell-to-cell
contact appears to alter their steroidogenic potential and morpho-
logic appearance in vitro. For these reasons the use of intact follicle
walls without separation of the thecal and granulosa compartments
may have to be considered as a test system if problems are
encountered with isolated cell systems.
Selection of the species for use depends primarily on availability
of adequate numbers of physiologically matured follicles or corpora
lutea. While diethylstilbestrol-treated immature rats can be used as a
source of ovarian cells (32), the numbers of cells are small and
require a substantial time investment for collection. Domestic
animals, by contrast, have much larger follicles, and the use of
slaughterhouse material of a polyovulatory species minimizes the
precollection time investment. Pigs and cows are the most desirable
large animals to use in this regard, and both have an extensive
literature available regarding their reproductive cycles, cell collection
techniques, and tissue culture. The cell system chosen should be an
easily exploitable model system in which known inhibitors of
steroidogenesis in both human and animal systems can be studied in
vitro to validate the animal model and the data extrapolated to
humans for more general application to quantitative risk assessment
of other suspicious compounds.
-------
33
.S
S
D
1
o
g"S>
•- S
*• M
tr'g
c B
o •*j
•| co
H O
o, c
S .2
B* .IS
§1
atures of Ovarian
Xenobiotic )
i
H
s a
r—t •*-> ^
E B S
2 S u
•M ^
00 M
•a a
<§ S
/*S
« o <3 j
OJ oo •*-* i i , •<: i
gas ' ' + z '
B| g
w < £
•**» §
O S) "« 1 + + 1 1
III
<: w £
C rS "M1
o 4,, -g S
S! S> •& !§ 'S 1 1 + + +
11111
BH W
bO .£H
"° n C
•§ S c & c S 3
2-a .8 "H .8 ™ o
gis's a'iaSo'i "i
•Soo C^Sci^'O^S .52
£"ai£ T3Ca-*ag-M <*-<
ia^(+_, agSo^'a*^^
<2 00 00 < (2 (S
o
3
s
a
0.
S"
o
c
i
I
Z
c
1
1 d .
.1 -1
tin O
.s S
•0 >.
1 §
o *u
**3 «H
*G cf
"o *^
C3 g>
c3 3
•S i
« "
"SB §
nus(-), absence
varying proportio
roportion.
itic digestion; org
•= _ cu £
S.S M E
« 2 .S N
1-2 l§
3 "3 g K
8 S S-§
B <> 3
2 2 o M
01 "c S a
S « "> 3
s «11
III!
•i- S ° S
S|1 |
•3-S fi.f
< H W Q
a -Q y ^3
-------
34
REFERENCES
1. Gellert, R. J., Bakke, J. L. and Lawrence, N. L.: Persistent estrus and altered
estogen sensitivity in rats treated neonatally with clomiphene citrate.
Fertil. Steril. 22: 224-250, 1971.
2. Clark, J. H. and McCormack, S. A.: The effect of clomid and other
triphenylethylene derivatives during pregnancy and the neonatal period. J.
Steroid Biochem. 12: 47, 1980.
S.Eroschenko, V. and Palmiter, R.: Estrogenicity of kepone in birds and
mammals. In: Estrogens in the Environment, J. McLachlan, Ed., Elsevier/
North Holland: New York; pp. 305-326, 1980.
4. Clark, J. H. and Peck, E. J., Jr.: Female Sex Steroids: Receptors and
Function: Springer-Verlag: Berlin; 245 pp., 1979.
S.Katzenellenbogen, J., Katzenellenbogen, B., Tatee, T., Robertson, D. and
Landratter, S.: The chemistry of estrogens and antiestrogens. In: Estrogens
in the Environment, J. McLachlan, Ed., Elsevier/North Holland: New
York; pp. 33, 1980.
6. Kupfer, D. and Bulger, W.: Estrogenic properties of DDT and its analogs. In:
Estrogens in the Environment, J. McLachlan, Ed., Elsevier/North Holland:
New York; pp. 239-263,1980.
7. Chan, L. and O'Malley, B. W.: Mechanism of action of sex steroid-hormones.
I. N. Engl. J. Med. 294: 1322-1328, 1976.
8. Chan, L. and O'Malley, B. W.: Mechanism of action of sex steroid-hormones.
II. N. Engl. J. Med. 294: 1372-1381, 1976.
9. Chan, L. and O'Malley, B. W.: Mechanism of action of sex steroid-hormones.
HI. N. Engl. J. Med. 294: 1430-1437,1976.
10. Clark, J. H., Anderson, J. N. and Peck, E. J., Jr.: Estrogen receptor
antiestrogen complex: atypical binding by uterine nuclei and effects on
uterine growth. Steroids 22: 707-713, 1973.
11. Bliss, C. I.: Statistical methods in vitamin research. In: Vitamin Methods,
P. Gyorgy, Ed., Academic Press: New York; pp. 448-609, 1951.
12. Pfaff, D. W.: Estrogens and Brain Function: Neural Analysis of a
Hormone-Controlled Mammalian Reproductive Behavior. Springer-Verlag:
New York; 272 pp., 1980.
IS.Gerall, A. A. and McCrady, R. E.: Receptivity scores of female rats
stimulated either manually or by males. J.Endrocrinol. 46: 55—59, 1970.
14. McClintock, M. K. and Adler, N. T.: The role of the female during
copulation in the wild and domestic Norway rat. Behavior 67(1—2):
67-96, 1978.
15. Mattison, D. R.: How xenobiotic compounds can destroy oocytes. Contemp.
Ob.Gyn. 15: 157-169,1980.
16. Mattison, D. R. and Thorgeirsspn, S. S.: Smoking and industrial pollution
and their effects on menopause and ovarian cancer. Lancet 1: 187—188,
1978.
17. Mattison, D. R.: Difference in sensitivity of rat and mouse primoridal
oocytes to destruction by polycyclic aromatic hydrocarbons. Chem. Biol.
Interact. 28: 133-137, 1979.
-------
35
18. Mattison, D. R. and Thorgeirsson, S. S.: Gonadal arylhydrocarbon hydroxy-
lase in rats and mice. Cancer Res. 38: 1368—1373, 1978.
19. Mattison, D. R. and Thorgeirsson, S. S.: Ovarian arylhydrocarbon hydroxy-
lase activity and primordial occyte toxicity of polycyclic aromatic hydro-
carbons in mice. Cancer Res. 39: 3471-3475, 1979.
20. Mattison, D. R., West, D. M. and Menard, R. A.: Differences in
benzo(a)pyrene metabolic profile in rat and mouse ovary. Biochem.
Pharmacol. 25: 2101-2104, 1979.
21. Zuckerman, S.: The Ovary. Academic Press: New York; 600 pp., 1962.
22.Pedersen, T. and Peters, H.: Proposal for a classification of oocytes and
follicles in the mouse ovary. J. Reprod. Fertil. 17: 555-557, 1968.
23. Armstrong, D. T. and Dorrington, J. H.: Estrogen biosynthesis in the ovaries
and testes. In: Regulatory Mechanisms Affecting Gonadal Hormone
Action, Advances in Sex Hormone Research, J.A. Thomas, and R. L.
Singhal, Eds., University Park Press: Baltimore; 3: 217-258, 1977.
24. Stouffer, R. L., Nixon, W. E. and Hodgen, G. E.: Estrogen inhibition of basal
and gonadotropin-stimulated progesterone production by Rhesus monkey
luteal cells in vitro. Endocrinol. 101: 1157-1163,1977.
25. Williams, M. T., Roth, M. S., Marsh, J.M. and LeMaire, W. J.: Inhibition of
Human chorionic gonadotropin-induced progesterone synthesis by estra-
diol in isolated human luteal cells. J. Clin. Endocrinol. Metab. 48:
437-440, 1979.
26. Goldberg, V. J. and Ramwell, P. W.: Role of prostaglandins in reproduction.
Physiol. Rev. 55: 325-351, 1975.
27. McNatty, K. P., Makris, A., DeGrazia, C., Osathanondh, R. and Ryan, K. J.:
The production of progesterone, androgens and estrogens by granulosa-
cells, thecal tissue, and stromal tissue from human ovaries, in vitro. J. Clin.
Endocrinol. Metab. 49: 687-699, 1979.
28. Haney, A. F. and Schomberg, D. W.: Steroidal modulation of progesterone
secretion by granulosa-cells from large porcine follicles: a role for
androgens and estrogens in controlling steroidogenesis. Biol. Reprod. 19:
242,1978.
29. Erickson, G. F., Wang, C. and Hsueh, A. J. W.: FSH induction of functional
LH receptors in granulosa cells cultured in a chemically defined medium.
Nature 279: .336-338, 1979.
30. Gulyas, B. J., Yuan, L. C. and Hodgen, G. D.: Progesterone production by
dispersed monkey (Macaca Mulatto) luteal cells after exposure to trypsin.
Steroids 35: 43-51, 1980.
31. Channing, C. P.: Influences of the in vivo and in vitro hormonal environment
upon luteinization of granulosa cells in tissue culture. Recent Prog. Horm.
Res. 26: 589-622, 1970.
32. Hall, P. F. and Young, D. G.: Site of action of trophic hormones upon the
biosynthetic pathways to steroid hormones. Endocrinol. 82: 559, 1968.
-------
-------
II. GLOSSARY OF TERMS USED IN
FEMALE REPRODUCTION
adenosis—a nonneoplastic glandular disease that occurs in the
uterine arnix and upper vagina.
amenorrhea—absence or abnormal cessation of the menses.
androgen—a class of steroid hormones produced in the gonads and
adrenal cortex that regulate masculine sexual characteristics; a
generic term for agents that encourage the development of or prevent
changes in male sex characteristics; a precursor of estrogens.
androgen antagonist or antiandrogen—agent that opposes or im-
pedes the action of an androgen.
anovulation—suspension or cessation of the escape of ova from the
follicles.
corpus luteum—an endocrine body formed in ovary at site of
ruptured Graafian follicle that. secretes an estrogenic and pro-
gestagenic hormone.
diestrus—quiescent period following ovulation in the estrous cycle
of female mammals in which the uterus prepares for reception of a
fertilized ovum.
dopamine or hydroxytyramine—an intermediate in tyrosine catabo-
lism and the precursor of norepinephrine and epinephrine.
ectopic pregnancy—pregnancy occurring outside the uterine cavity
egg—female sexual cell.
estradiol—an estrogenic hormone (C18H24O2) produced by follicle
cells of the vertebrate ovary; provokes estrus and proliferation of the
human endometrium,
estrogen—estrogenic hormone; generic term for various natural or
synthetic substances that produce estrus.
estrogen agonist—an agent that has a biological activity similar to
that of the physiological estrogens.
37
-------
38
estrogen antagonist or antiestrogen—agent that opposes or impedes
the action of an estrogen.
estrus—phase of the sexual cycle of female mammals characterized
by willingness to mate and in intact animals when ovulation occurs.
follicle (ovarian)—one of the vascular bodies in the ovary, con-
taining the oocytes.
follicle-stimulating hormone or FSH—a glycoprotein hormone
secreted by the anterior pituitary of vertebrates that promotes
spermatogenesis and stimulates growth and secretion of the Graafian
follicle.
galactorrhea—continued discharge of milk from the breasts in the
intervals between nurshing or after weaning.
gonadotropin—a substance that acts to stimulate the gonads.
Graafian follicle—mature mammalian ovum with its surrounding
epithelial cells.
s
gynecomastia—excessive development of the male mammary glands,
sometimes leading to milk secretion.
hypothalamic-pituitary-ovarian axis—the hormonal interactions that
link and control female reproduction.
hypothalamic-hypophyseal complex—the structural and hormonal
relationships between the hypothalamus and the pituitary.
lactation—the production of milk; the period following childbirth
during which milk is formed in the breasts.
luteinizing hormone or LH—glycoprotein hormone secreted by the
adenohypophysis of vertebrates that stimulates hormone production
by interstitial cells of gonads.
menopause—natural physiologic cessation of menustration, nor-
mally occurring in the last half of the fourth decade.
oligomenorrhea—prolongation of menstrual cycle beyond average
limits.
-------
39
oocyte—female ovarian germ cell present after birth.
ovarian cyclicity—the periodic changes observed in the ovary
associated with follicular growth, ovulation, and corpus luteum
function.
ovulation—discharge of an ovum or ovule from a Graafian folicle in
the ovary.
parturition—labor; giving birth.
-postpartum estrus—estrus with ovulation and corpus luteum pro-
duction which occurs in some species immediately after birth of
offspring.
progesterone—a steroid hormone (C21H3oO2) produced in corpus
luteum, placenta, testes, and adrenals that plays a physiological role
in the luteal phase of menstrual cycle and maintenance of pregnancy;
also an intermediate in biosynthesis of androgens and estrogens.
prolactin—a protein hormone produced by adenohypophysis that
stimulates secretion of milk and promotes functional activity of the
corpus luteum.
prostaglandins—various 20-carbon-atom compounds, formed from
essential fatty acids, that physiologically affect the female reproduc-
tive organs, the nervous system, and metabolism.
puberty—period at which the generative organs become capable of
reproduction.
relative binding affinity—the degree to which a ligand, compared to
standard ligand, is bound to a receptor.
secondary amenorrhea—any case in which the menses appeared at
puberty but have been suppressed.
steroidogenesis—enzymatic steps converting acetate and cholesterol
to sex steroids, glucosteroids, or mineralocorticoids.
testosterone—a biologically potent androgenic steroid which may be
released from the gonads and adrenal glands.
-------
40
virilization or masculinization—the assumption of male characteris-
tics by a female because of excessive production of androgenic
substances or masculinizing tumors of the ovaries.
-------
CHAPTER 3
CONSIDERATIONS IN EVALUATING RISK TO
MALE REPRODUCTION
INTRODUCTION
During evolution the reproductive patterns of mammals, in-
cluding man, were determined to a considerable extent by the nature
of the environment. Similarly today a variety of natural environ-
mental factors may alter reproductive activity and fertility. Potential
hazards to man's reproductive state are present in the environment as
pollutants whose effects are often not likely to be clear-cut. Thus,
sensitive assessment systems are needed. However, it is not clear how
the potential effects on male reproduction can best be assessed.
Detailed information about many aspects of male reproduction
exists, but there is little firsthand experience with detection in
animals of subtle effects of either new chemicals or environmental
hazards. The task of detecting effects that may be reflected only
marginally in fertility performance is made more difficult by the
variability of different reproductive parameters such as the concen-
tration of sperm in an ejaculate, the total number of sperm
ejaculated, and the sperm morphology within a population of
"normal" men hi our society. Hence, our understanding of the
reasons for this variation and our ability to evaluate subtle responses
to environmental hazards is minimal, and it is important to recognize
the embryonic state of our abilities in this regard.
It is clear that changes in human reproductive function induced
by environmental hazards might be reflected in reproductive be-
havior; in circulating levels of hormones such as follicle-stimulating
hormone (FSH), luteinizing hormone (LH), and testosterone; or in
testicular and epididymal function as evidenced by spermatogenic
activity, fertilizing potential of the ejaculate, and ability of the sperm
genome to support normal development after fertilization. Monitor-
ing of a human population for the normality of any of these
functions and assessment of the risk of certain levels of a chemical
hazard require objective criteria that are measurable hi man and,
when a new chemical is to be evaluated, in an appropriate animal
model. Ejaculates of human and of animal semen contain a
41
-------
42
considerable heterogeneity of spermatozoa, and several parameters of
ejaculates may vary considerably from sample to sample. Nonethe-
less, there is now a reasonable knowledge of many facets of the
normal physiology and of the variations to be expected for many
specific parameters in animal models and to a lesser extent for man.
The guidelines described below are based on current knowledge of
the physiology of male reproduction in mammals, including man,
and suggest an approach to risk assessment of existing and potential
chemical hazards for reproductive function.
Aspects of the Problem
The single most sensitive and important parameter for human
fertility is the total number of motile sperm in an ejaculate (1, 2). It
has not been possible to set an exact limit on the minimal number of
motile sperm per ejaculate, or what is more commonly reported as
concentration of sperm or semen, necessary for fertility in man. A
male ejaculating as few as 1 X 106 sperm per milliliter may prove
fertile occasionally (3), but in most cases low numbers of sperm per
ejaculate bear an obvious relationship to infertility. For example,
sperm concentrations below 10 million, of from 10 to 20 million,
and from 20 to 40 million per milliliter are associated with a risk of
infertility that is, respectively, tenfold, fivefold, and threefold higher
than for individuals with normal spermatozoal concentration, that is,
60 to 160 million per milliliter (4). Because another study (5) shows
that relative risks are fourfold and twofold higher for men with
sperm concentrations of below 10 million and between 10 and 20
million per milliliter, respectively, it is quite possible that a twofold
reduction in sperm concentration in individuals with sperm counts
below 40 million per milliliter will double the incidence of infertility.
Several characteristics of human semen and testicular function reflect
a low efficiency (2, 6). Human testes may function often at the
threshold of pathology (2, 7, 8) and may be particularly sensitive to
toxic agents compared with the testes of animals commonly used to
study testicular function.
The yield of spermatozoa from spermatogonia, the rate of sperm
production per gram of testis, and the percentage of morphologically
normal sperm in ejaculates are lowest in man among the many
mammals studied (2,6—8). The median number of sperm
(~200 X 106 per ejaculate) is only fourfold higher than the value
(50 X 106 per ejaculate) below which fertility becomes significantly
reduced (9). In contrast, the number of sperm in an ejaculate of bull
-------
43
semen (7 X 109), is 1400-fold higher than the value of 5 X 106
sufficient to achieve maximal fertility by artificial insemination (10),
and a smaller animal, the rabbit, also shows a large differential. It is
possible that a given set of conditions in the environment may cause
infertility in man more readily than in experimental animals. Several
agents, including radiation (11—14), chemotherapeutic drugs
(15—18), and dibromochloropropane (19—22), reduce motile sperm
concentrations and affect fertility.
Selection of an Animal Model
Evaluation of compounds for potential risk to human males
requires one or more animal models. The selection and use of these
models for testing end points that signify a reproductive hazard
generally is more specialized than that for most toxicology or
mutagenesis testing. The relevant end points depend on integrated
functional aspects that can be monitored with ease only in certain
species. The use of two species reduces the possibility of missing a
hazardous agent during testing.
Parallel testing of both rat and rabbit seems most suitable.
Although a number of laboratory or domesticated species might be
used, rabbits and rats offer several advantages in comparison to dogs
and subhuman primates. Rabbits have a high, predictable libido that
may be useful in assessing risks to sexual behavior. More importantly,
all sequential phases of the conception process (i.e., endocrine
function, spermatogenesis, sperm maturation, ejaculation, sperm
capacitation in the female, and fertilization) are easily evaluated,
quantified, and manipulated throughout the year. The ability to
characterize the whole ejaculate quantitatively and qualitatively and
the ability to collect the ejaculate with ease using an artificial vagina
make the rabbit a key test model for sensitive assessment of possible
harmful effects of environmental agents on male reproduction.
The rat is also a very useful model and is preferable to the mouse
or hamster because of the rat's widespread use in lexicological
research, the large base of knowledge of its reproductive processes,
its relatively low cost, its convenient size for weighing organs, and
the fact that it breeds readily under laboratory conditions. The rat is
less useful than the rabbit, because more of the measurements
require invasive procedures and/or sacrifice of the individual/ The
characteristics of several potential models are summarized in Table 7
(1, 23).
-------
44
li
S
1
M
"w
•g
1
•a
a
1
u.
.5
c
o
o
3
O
1
O
_o
a
3
w
M
*2
.g
4>
S
TABLE 7
•1
>» *w5*
"c i
O £
IS £.
2T
o|
T3
« J ^
3 o> 2
<5 a- ^
&
z.
rt
«
s
o
s
0
vd
"0
NO
CO
p.
o
ON
c*i
vo
CO*
tn
3
O
0>
a
.a
of cycle of sei
c
.2
COCSVOONVO CSOCO •<*•
v^^iOC-^ <=>^0^ ^JQS
rH TJ-
o\ o »o
o^o»ooot— T-HOCO
r^voc^«-ic^ O»-O>O ?— iO^J"O ^
•^h in en vo co O* t-J o 04 o O e--
^ -1 ^c^
CO O CO
COCOC"»-HCS OO^TJ"
^c--o'oi +-> rt
«j 3 -O
*8&
-------
45
"CJ
§
.s
s
o
CJ
«
3
H
S
03
5? ^
C cS
o .e
-s-
"> "a
O 5f
Q a>
•a
B
|| 3
a>
z
^"^
+jj
*
a>
o
McAcaco ^^eocnwsoswi
>.>->,>, 2Z>.>,>,>->.
C) ^J
cACAcAco wsertwsvswa
si Sj sf ^ ^izsisJ1^^!^
c« -co co co waOTenwawa
^ ^ ^ ^ ^^^^^
ssss sssssss
>,>.>.>. >H >H
SSSo SjOeSooSo
[>•( ^X )>« £5 J>4 ^ J>-C ^ IZ] ^M J>l
z||z ^z^zz^z
1.1° i
w> P S 3 O «
S S <2 o -c ^
I -a S c .fe " S
1 S « -2 SS>-«
o 1 || | ^-~ig
|lll||j|i.|lll
c
3
_O oi
3. a
S "3
it 3
Q |
1 'i
x'E e .
O o O ^
rt -^3 fS
ro ^ ^
II ,2 3 G
a, o o a
fi «J i
-------
46
Tests for Evaluating Reproductive Damage
We reviewed a wide spectrum of test systems. Table 8 (24—56)
lists tests that were considered to be suitable for qualitative and
quantitative risk assessment (for detailed consideration, see the
Appendix to this chapter). Table 9 (including Refs. 57-70) lists
additional tests that were considered but rejected because they are
insensitive, redundant, not cost effective, too difficult to perform
unless within a research setting, not validated, too controversial, or in
need of further development.
The tests in Table 8 evaluate the endocrine control of male
reproduction and the number and quality of sperm produced, or
they measure fertility. Since fertility is related to the number of
normal spermatozoa in the ejaculate, the analyses of seminal quality
are indirect measures of fertility. However, comprehensive seminal
analysis is a much more sensitive end point for detection of a toxic
effect than a breeding experiment using natural mating, because in
experimental animals the number of sperm ejaculated greatly exceeds
the number necessary for fertility (10,71). The various tests are
stratified into a sequence (tests 1-3, S, E) ranging from a prelimi-
nary screen to more detailed studies.
Coefficients of variation (CV) might be used to determine the
sensitivity of a study that compares treated animals with controls
(see Appendix). Generally, measurements of testicular and epididy-
mal sperm numbers give highly reproducible values in control animals
with coefficients of variation between animals of 15% or less (24).
This degree of reproducibility ensures that the test will be quite
sensitive, even with relatively few animals (Table 10). Although there
is appreciable variation (CV = 70%) in sperm concentration or total
number of sperm per ejaculate in semen collected, from rabbits or
bulls (72, 73), this variation can be reduced somewhat by using a
uniform interval between seminal collections and standardized
procedures in chronic studies. Sperm motility and morphology are
much more constant than sperm number or concentration, especially
within individuals of a species (10, 74). These two assays usually are
performed in a subjective manner, and efforts must be made to
minimize this subjectivity (2). In some species motility can be
evaluated objectively by measurements made on time exposure
negatives (75) or probably better on videotape recordings (49).
Morphology of spermatozoa from individuals always should be
compared to control samples analyzed concurrently by the same
observer. The slides and/or videotapes should be retained for
validation by an outside observer.
-------
47
TABLE 8 Tests Considered Useful for Screening Toxic Compounds"
Test
Body Weight
Testis
Size in situ
Weight
Spermatid reserves
Gross histology
Nonfunctional tubules (%)
Tubules with lumen sperm (%)
Tubule diameter
Counts of leptotene spermatocytes
Epididymis
Weight of distal half
Number of sperm in distal half
Motility of sperm, distal end (%)
Gross sperm morphology, distal end (%)
Detailed sperm morphology, distal end (%)
Gross histology
Accessory Sex Glands
Weight of vesicular glands
Weight of total accessory sex glands
Semen
Total volume
Gel-free volume
Sperm concentration
Total sperm/ejaculate
Total sperm/day of abstinence
Sperm motility, visual (%)
Sperm motility, videotape (% and velocity)
Gross sperm morphology
Detailed sperm morphology
Concentration of agent in sperm
Concentration of agent in seminal plasma
Concentration of agent in blood
Biochemical analyses of sperm/seminal plasma
Endocrine
Luteinizing hormone
Follicle-stimulating hormone
Testosterone
Gonadotropin-releasing hormone
Fertility
Ratio exposed: pregnant females
Number embryos or young per pregnant female
Ratio viable embryos: corpora lutea
Ratio implantation: corpora lutea
Number 2-8 cell eggs
Number unfertilized eggs
Number abnormal eggs
Sperm per ovum
Number of corpora lutea
In Vitro
Incubation of sperm in agent
Hamster egg penetration test
Rat
1-3
1-3
1-3
1-3*
1-3C
2,3*
2,3
2,3
1-3
1-3
1-3
1-3
1
2,3
NA
1-3
NA
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
2,3
2,3
2,3
2,3
1-3
1-3
1-3
1-3
3*
3«
3«
3«
3«
NA
NA
Rabbit
1-3
1-3
1-3
1-3*
1-3^
2,3*
2,3
2,3
1-3
1-3
1-3
1-3
1
2,3
NA
NA
i-3
1-3
1-3
1-3
1-3
1-3
1-3
2,3
1
.2,3
NA
3d
3d
NA
2,3
2,3
2,3
2,3
1-3
1-3
NA
NA
NA
NA
NA
NA
NA
3f
NA
Human
S,E
S,E
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
NP
E
NA
E
E
E
E
E
NA
E
NA
Ed
Ed
NA
E
S,E
E
E
NP
NP
NP
NP
NP
NP
NP
NP
NP '
Ef
E
Reference
24-26
24-27
24,25,27-30
26, 31-34
13, 18, 35
26, 32, 33
26
32, 34
25,27
27, 36, 37
38-40
41,42
6,41,42
26,27,43
26,27,43
2, 24, 26, 44-46
24, 26, 44-46
2, 24, 26, 44-46
24, 26, 44-48
2, 24, 26, 48
2,26
2,49
2
2,42,50
2,46,51
52,53
52,53
52,53
52,53
54
54
54
54
55
55
55
55
56
-------
48
TABLE 8 (Continued)
"Test 1 - initial at maximum tolerated dose (MTD), or MTD and 0.7 MTD, run for exactly six
cycles of the seminiferous epithelium. A similarly significant change (probably >15%) in any criterion
would be evidence of an effect.
Test 2 = dose response at MTD, at -1 and —2 log dose, and down to human level if known or until
no response is obtained in any test; run for exactly six cycles of the seminiferous epithelium.
Test 3 " long term, reversibility; several doses and time periods. Expose to at least three doses for
at least 6 cycles of the seminiferous epithelium (kill 1/3 of males) and then allow recovery for 6 cycles
(kiU 1/3 of males) and 12 cycles (kill 1/3 of males). Recovery at 12 cycles after termination of
treatment should be to at least 90% of control level to show complete reversibility.
S = procedure useful for screening humans in industrial setting.
E = procedure useful for studying individuals thought to be exposed to an agent. Evaluation of
human semen should use samples obtained after 2 to 5 days of abstinence with samples taken over
time.
NA » not necessary.
NP = not practical or possible.
^Especially important when studying recovery.
cSave tissue from level 1 test, fix in Bouins, for possible later use.
''in 3 samples taken near end of treatment and then in additional samples to get clearance rate.
^Female rats killed 18-24 hours after mating to evaluate fertility, sperm penetrating ability, and
sperm transport.
/If compound is detected in seminal plasma of rabbit or man, incubate both rabbit sperm and
human sperm from normal donors and determine a dose response of sperm to the drug in vitro.
Evaluate percentage of motile sperm over time at 37° or in vitro penetration of hamster oocyte.
All of the tests listed in Table 8 are feasible in most well-
equipped laboratories. The phase-contrast microscope and video
micrography equipment (estimated additional cost: $6000) are the
only nonstandard requirements. The training period necessary to
conduct the tests in an accurate and precise manner is not excessive.
The tests selected can be used to (a) detect an effect of a test
compound on male reproduction and (b) serve as a basis for
estimating an acceptable level of exposure.
The tests listed will yield quantitative data that are amenable to
efficient statistical analyses and that have a sufficient range of values
to enable establishment of dose-response curves. The variability of
the tests is shown in Table 10 for most parameters.
The proposed tests are for the most part quite specific for
reproductive toxicity. Results of each test should not be affected by
other body systems or, except for a possible decrease in testosterone
level, cause changes in other aspects of body function.
Any subchronic or chronic test used to evaluate effects of an
agent on male reproduction must extend over 6 cycles of the
seminiferous epithelium, when it is assumed that an agent bio-
accumulates to a steady state within 1 cycle (23). This interval is
based on (a) the time needed to reach a steady state concentration of
-------
49
TABLE 9 Reasons for Rejection of Potential Evaluation Tests
Considered by Male Reproductive Subgroup
Test
References
Reasons for rejection"
Tonometric measurement of testicular consistency 26,57
Qualitative testicular histology 24, 31, 34
Stage of cycle at which spermiation occurs 24, 31, 34
Quantitative testicular histology
Counts of degenerating germ cells 18, 35,58
Complete germ cell counts 18, 35,58
Stem cell counts 18,35,58
Relative frequency of stages of cycle 18,35,58
Epididymal histology 59, 60
Biochemistry of epididymal fluids 2,61
Histology of accessory sex glands 62
Biochemical analysis of sperm 2
Sperm membrane characteristics 63, 64
Biochemical analysis of seminal plasma 65
Evaluation of sperm metabolism 65,66
Fluorescent Y bodies in spermatozoa 67, 68
Flow cytometry of spermatozoa 69
Karyotyping human sperm pronuclei 70
Cervical mucous penetration test
Studies on prepuberal animals
US (rat), NV (rabbit), FR (human)
I
RR,UR
UR
UR,$$
UR,$$,FR (rabbit)
UR,I
I,NR
NR
NR
NR,$$
NV.FR
NR
NR,$$
NV.NR
UR, NV, FR
FR
UR (human)
NR (rat, rabbit)6
CUS = unsuitable for species
NV = not validated
RR - redundant
I = insensitive
NR = not relevant
UR = only in specialist lab
FR = future research
$$ = too costly
"Studies on animals treated prior to puberty have not been included for the following reasons. It
would involve a redetermination of maximum-tolerated-dose levels for young, growing animals. The
choice of age period of exposure is a complex topic and sufficient time was not available to adequately
consider this. Humans are exposed to many of the agents that would cause reproductive problems
primarily through occupational exposure after puberty. Some agents (radiation, cyclophosphamide)
that cause reproductive problems with prepubeial exposure also affect postpuberal males. Nonetheless,
unique developmental processes occur in testicular development prior to and during puberty, and
therefore a possibility exists that some agents would only affect the prepuberal male. The group of
tests proposed in Table 8 would provide a sensitive measure of such effects, if animals exposed at any
time during puberty or throughout their development were analyzed after reaching sexual maturity.
agent in the target organs of the rabbit or rat, (b) the concepts that
an agent acting directly or indirectly on the germinal epithelium may
act on a specific type of cell and that affected germ cells may
develop for some time before they degenerate, (c) the fact that
damage to germ cells is most evident by absence of certain types of
germ cells, and (d) qualitative change in germ cells may not be
readily discernible until active spermatozoa pass into the cauda
epididymidis or ejaculated semen. The present protocol assumes that
attainment of a steady state concentration of an agent requires an
interval equal to one cycle of the seminiferous epithelium. Forma-
tion of primary spermatocytes from renewing spermatogonia in the
-------
50
TABLE 10 Approximate Variation Between Animals for Suggested Test
Criteria (CV)fl Coefficient of Variation (%)
Rabbit model0
Criterion
Body weight
Testis
Weight
Size in situ
Spermatid reserves per testis
Spermatid reserves per gram
Tubule diameter
Epididymis
Weight of distal half
Number of sperm in distal half
Motility of sperm, distal end (%)
Gross sperm morphology, distal end (%)
Detailed sperm morphology, distal end (%)
Accessory sex glands
Weight vesicular glands
Weight total accessory sex glands
Semen
Total volume
Gel-free volume
Sperm concentration
Total sperm per ejaculate
Total sperm per day of abstinence
Sperm motility, visual (%)
Sperm motility, videotape (%)
Gross sperm morphology
Detailed sperm morphology
Concentration of agent in seminal plasma
Concentration of agent in blood
Endocrine
Luteinizing hormone
Follicle-stimulating hormone
Testosterone
Gonadotrophin-releasing-hormone stimulation
Fertility^
Ratio of exposed to pregnant females
No. embryos or young per exposed female
No. embryos or young per pregnant female
Ratio of embryos to corpora lutea
Rat model6
(Wistar)
20
5
NAa
11
8
_e
13
20
-
-
-
26
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
80
65
33
-
"Data are not available to allow calculation of sensitivity of the
*Ref. 27.
cRefs.25,37,and44.
j j '
NA = not applicable.
e— = data not available.
•fipor controls: with treatment, variability may
be greater.
(New Zealand
White)
37
18
9
24
9
5
13
52
-
-
-
10
-------
51
rat requires about 1.5 cycles, and spermiation occurs about 3 cycles
of the seminiferous epithelium after those sperm have become
primary spermatocytes. Passage of sperm through the epididymis
into the distal cauda or ejaculated semen requires 1.0 to 1.5 cycles,
depending on the species and frequency of ejaculation. Conse-
quently, if an agent acted on A-type spermatogonia, a decrease in
number of sperm ejaculated or in the fertility of sperm from the
cauda epididymidis might not occur for 5 to 6 cycles
(1.5 + 3.0 + 1.0 = 5.5 cycles) of the seminiferous epithelium. If the
agent resulted in degeneration of pachytene spermatocytes, an
alteration in semen characteristics or fertility might be expected to
occur after 4.0 to 4.5 cycles. However, with continuous exposure to
the test compound, such a lesion would remain detectable in the
semen or by examination of testicular histology at the end of 6
cycles of the seminiferous epithelium.
By testing male rats or rabbits for their fertility after 5 cycles, a
depression in fertility caused by a compound inducing a qualitative
change in sperm function should be detectable, since this probably
would affect spermatocytes or spermatids. Allowing 6 to 8 days of
sexual rest between the end of fertility testing and necropsy of test
males after 6 cycles provides time for restoration of the normal
population of sperm in the cauda epididymidis hi males receiving
doses that do not suppress daily sperm production. If sperm
production is low in test males, the reserve level in the cauda will
reflect this,, but sufficient sperm may still be present to allow
assessment of sperm motility and morphology.
For these reasons, evaluation of an agent, administered chroni-
cally, for effects on male reproduction should include fertility tests
after 5 cycles and examination of the testes, epididymides, accessory
sex glands, and plasma hormone levels after 6 cycles of the
seminiferous epithelium.
If an agent is shown to alter male reproduction in a test
extending over 6 cycles of chronic exposure (Test 1 or Test 2 as
described under protocols for testing), it may be desirable to
determine if the effect is reversible (Test 3). A test of reversibility
should extend over 18 cycles of the seminiferous epithelium. Chronic
exposure to the agent should extend over 6 cycles, and 12 cycles
should elapse after the termination of exposure to allow for
restoration of normal reproductive function.
Although the recovery period in man is usually longer, these
animal data should provide a clear indication if complete recovery
will occur in man (compare radiation data of Meistrich et al. [13] in
the mouse and Rowley et al. [ 14] in man).
-------
52
Evaluation of Reproductive Damage in Exposed or
Potentially Exposed Men
General
Two types of studies in humans seem particularly relevant to the
objectives of a male reproduction risk assessment. The first involves
surveillance studies, in which periodic checking is done on men in a
setting (e.g., industrial or agricultural) that might, in the future,
involve the risk of a reproductive defect. An example of such a
setting would be a chemical company in which substances are
prepared that are known from animal studies to cause reproductive
toxicity when administered in high dosage but not at a dosage up to
10 times that expected for human exposure. The safety factor may
be variable, however, and depends on the quality of the animal study
from which it was derived. This type of surveillance is important for
at least two reasons: (a) the sensitivity of men may be greater than
allowed for by the tenfold safety factor and (b) the exposure of the
workers may be greater than originally estimated. The methods for
surveillance of this type could be quite innocuous and could be
incorporated into an annual medical checkup if this were already a
practice.
The second type of study would be of men who have been or are
being exposed to a known reproductive toxin in dosages likely to be
toxic in man as based on animal studies. This type of study could be
used in men exposed to high dosages of one or more general toxins
for which the effect on testicular function had not been studied
carefully in animals.
Surveillance studies
Men could be asked yearly whether they have been attempting to
cause a pregnancy and have been unable to do so. The prevalence of
infertility in couples within the reproductive age group is approxi-
mately 15% (74). If more than 20% of men between 19 and 35 have
been unable to produce a pregnancy in over one year of unprotected
intercourse, a possible toxic effect should be looked for in a rigorous
manner as outlined in the following section on known toxic
exposure.
Testicular length could be measured on annual physical exami-
nation. If the distribution of testicular size for men falls significantly
below the lower norm (3.5 cm for Caucasian and Black) for that age
group and ethnic background (76), a toxic effect should be
suspected.
-------
.53
Blood levels of follicle-stimulating hormone could be measured
yearly. If mean levels significantly vary from those of age-matched
controls, a toxic effect should be suspected.
If any of these three variables suggest testicular toxicity, a more
detailed study of the population should be undertaken, as outlined
below.
Study of men with known toxic exposure
Where a human population is suspected of being at reproductive
risk because of environmental hazards, a number of potentially toxic
agents may be involved, and the duration and level of exposure may
vary within the population. Although each potential toxicant should
be carefully tested by the laboratory screening methods outlined in
this document, it would be useful to make a more immediate and
direct assessment of fertility potential in the exposed population.
The requirements for this include capacity for rapid response to the
subjects, feasibility, sensitivity, and data that can be analyzed
statistically. The data obtained in such studies should provide an
initial indication of the degree of testicular damage, and where an
environmental reproductive hazard has been identified, more detailed
studies may be undertaken to characterize objectively male reproduc-
tive dysfunction.
To carry out these studies, a specialized team and a modest
amount of equipment would be needed. The latter could be installed
at locally available facilities, or a mobile laboratory could be
equipped. Detailed medical, reproductive, and occupational histories
should be taken from each exposed subject and a physical examina-
tion given. At least five semen samples should be evaluated per
individual at two-day intervals. Objective data on testicular size and
consistency could be obtained by sonography and tonometry. Blood
and urine could be obtained at this time for endocrine studies and/or
toxicant levels. Controls to be studied must be carefully chosen and
matched. Before a national data base is established, individuals
should be selected according to epidemiological advice. For the
details of these analyses, see "Study of men with known toxic
exposure" in the Appendix to this chapter.
Statistically significant differences between the exposed and the
control groups (matched for age, occupation, geographical location)
in seminal fluid and blood hormone measurements would be
evidence for an effect of the exposure on male reproductive
function. An adverse effect would be expected to decrease sperm
counts, motilities, and numbers of sperm with normal morphology; if
-------
54
the effect were sufficiently severe, blood testosterone levels would
decrease. If the toxic effects were directly on the testis (as is the case
with the great majority of known toxins), follicle-stimulating
hormone (FSH) levels would increase. With a mild toxic effect on
the testis, blood FSH levels after administration of gonadotropin-
releasing hormone (GnRH) might exceed normal responses, even
when basal FSH levels are normal. If the toxic effect were primarily
on the pituitary gland or central nervous system, luteinizing hormone
(LH) and FSH levels would tend to decrease.
If not established initially (Surveillance studies, v. sup.), other
comparisons between the exposed and control groups should include
(a) rate of infertility as indicated by the number of men who have
not been able to induce a pregnancy in over one year of intercourse
without using contraceptives, and (b) testicular size, with particular
attention to the number of men with testicular length less than
3.5 cm.
Differences between the exposed and control groups in these last
two assessments, suggested also for the initial screen, will be found
when reproductive toxicity -is sufficiently severe. However, measure-
ments of fertility and testicular size would be expected to be less
sensitive in revealing mild defects in gonadal function than the
seminal fluid and blood hormone measurements described above.
Additional comment on human testing procedures
Blood samples for hormone measurement and noninvasive
procedures such as testicular length may be the more feasible
parameters to evaluate because of the added difficulty in obtaining
semen samples, in some human populations at least. However, where
these give equivocal results, it is likely that semen analyses will help
to resolve the fundamental dilemma.
An elevated FSH level is a sensitive indicator of decreased
function of the germinal epithelium in man and experimental
animals. However, while there is no doubt that increased FSH levels
usually imply decreased sperm production (9, 14, 21, 77, 78),
measuring FSH levels is probably a less dependable test than direct
sperm counts, for it is a consistent marker only of severe oli-
gozoospermia or azoospermia (53, 79, 80). Measurements of FSH
seem useful adjuncts to sperm counts, therefore, and indicators of
the direct action of toxic agents on pituitary function. Despite their
wide variability in man, total sperm per ejaculate are usually a more
sensitive measure of testicular damage than elevated FSH levels.
-------
55
Comprehensive semen analysis requires an assessment of sperm
morphology as well as total sperm per ejaculate. This aspect of the
human ejaculate has received considerable general comment as a
parameter that also often falls below the standards that might be
expected for animals living in the same area. Since about 30% of the
spermatozoa are abnormal in semen from a presumed fertile group of
men, only by using large groups of 100 or so persons would it be
possible to detect increases in abnormal spermatozoa of the order of
10% in cross-sectional studies in which only one to four samples are
collected for each man in exposed and control groups. Despite the
relatively variable morphology of a significant proportion of sperma-
tozoa in the human ejaculate, sperm morphology tends to be fairly
constant for one individual (81). This justifies the use of fewer men
in longitudinal studies where such studies can be undertaken (as
compared to postexposure analyses), since the men can then act as
their own controls (81).
Human sperm morphology classification is currently subjective,
personality oriented, and nonstandard (82). However, detailed "type
classification" (i.e., oval versus tapering versus amorphous) may not
be required to identify groups of individuals at reproductive risk. In
normal fertile human semen, sperm morphology is relatively uni-
form, more than 50% of the sperm having the typical "oval" shape.
In contrast, infertile human semen is characterized by a diversity of
abnormal sperm sizes and shapes. If objective, morphometric data
describing sperm size and shape (e.g., head length, width, area, and
circumference; tail midpiece width) were obtained from individuals
at potential risk, these could be compared statistically with data
from the matched control group. Significantly greater dispersion in
the morphometric parameters of the exposed group might indicate
increased reproductive dysfunction in the population. The magnitude
of differences between the exposed group and the control group
might also provide an indication of the severity of testicular damage.
As noted earlier, the methods for automatic evaluation of sperm
morphology are not well established and need considerable refine-
ment and validation (6).
Assessment of risk to men
Assessment of risk to reproductive performance and fertility in
men is inadequately tested at present. The quantitative assessment of
risk to general human health from exposure to environmental
toxicants has been approached by relating the probable or estimated
dose of a suspect toxic agent to the occurrence of deleterious effects
-------
56
on the basis of either epidemiologic data on human populations or of
experimental data from animal studies. It seems likely that threshold
effects will appear for most agents. Ideally, assessment of the effects
of these agents upon male reproductive function should be based
upon human epidemiologic data. However, there are few epidemio-
logic risk assessment data regarding the effect of environmental
agents on the fertility of men. Thus, at present a quantitative risk
assessment must depend on extrapolation to man of measurements
of the reproductive end points in experimental animal systems
discussed here. One approach to risk assessment estimates the
acceptable daily intake (ADI) of a chemical, defined as the exposure
level that is anticipated to be without risk to the species. It should be
cautioned that the ADI represents only a judgment, is not an
estimate of risk nor a guarantee of absolute safety, and is subject to
modification as additional relevant information becomes available.
To account for the uncertainties involved in extrapolating from
animals to man, the ADI includes an uncertainty or safety factor to
the highest no-adverse-effect level measured in an animal study. A
no-adverse-effect level is defined here as a dose for which no
significant difference is found between control and treated animals
for any of the end points measured adequately. It is important that a
statistically significant effect also be biologically significant. This
uncertainly factor will depend on (a) the animal species/strain; (b)
the quality of the experimental data; (c) the availability of
comparative pharmacokinetic information on the animal species' and
man's absorption, distribution, metabolism, binding, and elimination
of the chemical; and (d) any other relevant comparative information
on structurally similar chemicals. In the absence of these comparative
data, we should follow the guidelines of the Safe Drinking Water
Committee, National Research Council of the National Academy of
Sciences (83), and recommend an arbitrary uncertainty factor of 100
for adequate animal studies. In the case of human male reproduction,
the size of this factor seems more than justified by increasing
evidence that the human testis functions less efficiently and possibly
closer to a point of pathology than that of the animal models
recommended (2, 7, 8). Thus, for an agent causing a reversible action
in a model animal, the ADI would be 0.01 times the no-adverse-
effect level for the most sensitive criterion and the most sensitive
species evaluated, whether rat or rabbit. A daily exposure or intake
above this level represents a risk of reproductive damage to human
males. For irreversible effects on male reproductive function, we feel
we can make no recommendation for a quantitative risk assessment.
-------
57
Protocols for Testing Compounds with Animal Models
The actual criteria to be evaluated in each test are shown in Table
8. The time schedules for conducting Tests 1, 2, and 3 are shown in
Table 11. If pharmacological studies show that the test compound
may bioaccumulate so that the body burden increases beyond an
interval equal to one cycle of the seminiferous epithelium, the
treatment interval of both Test 1 and Test 2 must be increased
appropriately. At least 5 cycles should elapse after reaching
maximum body load.
Test 1 — initial screen
As an initial screening procedure, animal exposure will be greater
than or equal to half of a maximum tolerated dose (>0.5 MTD) of
the test agent for an interval equal to 6 cycles of the germinal
epithelium. An initial screen using an acute exposure is considered to
be unnecessary, because the subchronic test is more sensitive.
To initially assess risk to male reproduction, a compound should
be subjected to in vivo tests utilizing both rats and rabbits. A
compound producing no statistically significant alteration in any
criterion for either species when given at >0.5 MTD would be
considered to be safe for humans (see safety factor in risk
assessment). A statistically significant alteration in any criterion
would necessitate conduct of a Test-2 evaluation to establish a
dose-response curve, unless manufacture or use of the agent were to
cease, or if a larger safety factor were used. Test 1 (Tables 8 and 11)
uses both rats and rabbits and is detailed in the Appendix to this
chapter. The fixed time schedule is designed to maximize the
probability of detecting any decrease in reproductive function.
Test 2 — dose response curve
1. The general approach used in the initial screen (Test 1) will be
used except that additional criteria of reproductive damage are
included (Table 7). The dose-response curve will include at least
three points, usually the dose used in Test 1, and —1 and —2 log
doses and must extend down to the human exposure level (if known)
or until no statistically significant response is obtained in any test.
Both 0-dose and untreated controls could be included to detect
effects of handling that might be associated with agent administra-
tion. If necessary, additional tests will be run to attain these end
points. Both rats and rabbits must be used. The fixed time schedule
(Table 11) is essential to measure accurately the extent of damage to
-------
58
TABLE 11 Chronology of Conduct for Test with Animal Models
(Expressed as Day of Study)
Test
Rat
Rabbit
Test 1 or 2a
Condition males
Obtain preexperimental body weight
Evaluate preexperimental semen
Initiate compound administration
Continue compound administration
Weigh weekly
Collect experimental semen (each 3-4 days)
Measure testis size weekly
Expose to females or artifically
inseminate females
Sexually rest males
Kill males
Kill females
Allow females to kindle
Test la
Condition males
Obtain preexperimental body weight
Evaluate preexperimental semen
Initiate compound administration
Continue compound administration
Collect experimental semen
Expose to females or artifically
inseminate females
Kill 1/3 of males
1/3 of males
1/3 of males
Kill females 18—24 hours past mating
12—18 days past mating
Allow females to kindle
-21-0
-14 + 0
NA6
day 0
0-77
0-78
NA
NA
65-71
71-78
78
83 - 89C
NA
-21-0
-7 + 0
NA
0
0-77
NA
NA
NA
215-221
78
155
232
216-222
233
NA
-28-0
-14 + 0
-14-0
day 0
0-64
0-65
3-4-53-54
0-65
54-57
58-65
65
NA
85-90
-28-0
-7 + 0
-14-0
0
0-64
35-64
118- 140
158-194
184-187
64
128
193
NA
NA
215-218
"Rats will be weighed weekly. Rabbits will be weighed weekly, testis size
measured weekly, and semen will be collected twice weekly (every 3 to 4 days).
Schedule is for a compound that does not accumulate for a long time; steady
state level in body tissues reached in <10—12 days.
°NA = not applicable.
cKill females 18 days after mating, as determined by a vaginal smear.
the different aspects of male reproductive function and to enable a
prediction of human risk.
2. In evaluating testicular histology, sections representing at least
two loci will be used. The diameter of 50 tubules will be measured;
the percentage of seminiferous tubule cross-sections (N = 250) having
mature spermatids lining the tubule lumen and the percentage of
-------
59
tubules (N = 250) devoid of germ cells other than spermatogonia will
be determined. Evaluations of the morphology of sperm in the cauda
epididymidis and ejaculated semen will be more comprehensive than
in Test 1. The serum concentrations of luteinizing hormone (LH),
FSH, and testosterone also will be determined.
Test 3 — recovery study
1. Test 3 is a long-term study designed to test the reversibility of
damage to male reproduction and also to evaluate sperm transport,
penetration of sperm into ova, and early embryonic death. Many
agents that cause degeneration of the germinal epithelium and
azoospermia will not damage the stem spermatogonia. If the latter
remain, eventual recovery of the germinal epithelium is likely. This
test measures recovery of the germinal epithelium and fertility, at 6
cycles and 12 cycles (155 days for rats and 128 days for rabbits)
after ending a 6-cycle exposure to the test compound. Although
recovery of the germinal epithelium might not be complete by 12
cycles after exposure, some onset of recovery probably should be
detectable by then if it will occur eventually. If the test compound is
one known to bioaccumulate, longer treatment periods (as used in
Test 2) and recovery periods (at least twice the duration of the
treatment period) are essential.
2. The criteria evaluated (Tables 8 and 11) are the same as those
in Test 2, except that data on fertilized rat eggs are necessary.
Consequently, each male will be exposed to four female rats. Two
females will be killed 18 to 24 hours after mating (as determined by
the presence of vaginal plugs) and ova recovered by flushing. The two
other female rats will be killed 12 to 18 days after mating.
Measurement of concentrations of the compound in blood and
seminal plasma at steady state are desirable, since these data may be
useful in predicting potential damage in humans and the prognosis
for recovery from such damage.
3. The time schedule (Table 11) for conduct of the study could
be modified by extending the treatment beyond 6 cycles, but the
timing of evaluations between days 215 and 233 for rats and days
184 and 218 for rabbits may not need to be altered.
4. Complete reversibility is considered to be restoration, to at
least 90% of control levels at 12 cycles after cessation of exposure, of
all criteria adversely altered in males after 6 cycles of exposure (in
Test 2 or Test 3) at a given dose.
-------
60
Research Needed
Until about 15 years ago only outline information about the
male tract was available (84—86). Although much precise data has
appeared since then for the animal models suggested and even some
for man (65, 87, 88), it is difficult to compare the two. Research
into this and related aspects as suggested below is a critical element
for establishment of a reliable assay and evaluation of risk in males.
1. Variance components for characteristics of semen (total
volume, total sperm per ejaculate, percentage of motile sperm, and
incidence of sperm abnormalities) are available for rabbits (see
Table 10) but have not been reported for man. This information
should be obtained for men of different age groups (<20, 21-30,
31_40, >40) with different life styles or occupations, so that
efficient and meaningful evaluations can be made.
2. The influence of an abstinence interval on characteristics of
human semen should be evaluated critically for men of 20 to 30, 30
to 40, and >40 years of age. Procedures for reducing the influence of
an abstinence interval on estimates of sperm production (e.g.,
normalization of data for each ejaculate by dividing by the number
of days of abstinence) should be evaluated. It is also unclear what
effect repeated ejaculation has on the absolute concentrations of
many seminal compounds that might be measured as indicators of
the activity of accessory glands in man or the amount of the test
agent in seminal plasma.
3. For the human, the relationship should be determined among
testicular size, tonometric measurements or testicular consistency,
and sperm output as well as other ejaculate characteristics.
4. Relationships need to be established among testicular histol-
ogy, ejaculate characteristics, sperm morphology, and fertility of
humans, rabbits, and rats. Indices of fertility should be calculated.
5. Automatic or semiautomatic morphometric procedures should
be further developed for analysis of the morphology of human
spermatozoa and spermatozoa from test animals. First-generation
systems for automated evaluations are available (e.g., at Lawrence
Livermore Laboratory), but the instrumentation and software need
additional refinement and validation before these techniques can be
applied routinely in analyses of human sperm morphology (89, 90).
Sperm morphometry can be obtained for living sperm cells or
from stained seminal smears, by methods becoming increasingly
automated, either using, flow cytometry (89) or tracing sperm shape
from the screen of a video monitor using an electronic planimeter-
digitizer integrated into a minicomputer (49). The use of sperm
-------
61
morphology for diagnosis or prognostication of specific reproductive
disorders will require a "type classification" of individual sperm
abnormalities. The determination of such a classification can be
made by computer on the basis of the morphometric data obtained.
Research should be encouraged to develop such computer software.
A classification system based on morphometric standards should be
developed.
6. A data bank should be established for (a) the control data
from screening tests with animals to build a large base for
computation of variations associated with each characteristic among
trials, locations, season, year, etc.; (b) the chemical nature of
compounds tested and found in an initial screen to damage some
aspect of male reproduction or to have no effect; and (c) the
chemical nature of all compounds found to have a deleterious effect
on male reproduction, the nature, extent, and incidence of damage in
each exposure dose, and interval to recovery.
7. Available data on effects of agents known to alter human male
reproduction should be correlated with data on their action in test
animals in a battery of tests. The repeatability and relative sensitivity
of the tests within and between species should be determined.
Recommendations for specific studies are as follows:
(a) Obtain more extensive and accurate analytical data on semen
from men exposed to chemotherapy and the effects of parallel levels
of chemotherapy in animals.
(b) Bring together existing radiation studies in human and
experimental animals for development of models in which to base
chemical risk assessment data.
(c) Obtain better data in experimental animals on effects of
dibromochloropropane or other agents known to be harmful to man.
8. The relative usefulness of basal FSH concentration in blood or
of FSH response to GnRH as indicators of testicular damage should
be compared with seminal analyses to determine their sensitivity (see
Table 10 and Appendix to this chapter). For screening of large
numbers of human males, it would be useful to know the single most
sensitive index of testicular toxicity. In man, a test based on a blood
sample is more practical than one requiring submission of a seminal
fluid sample.
9. The responsiveness of Leydig cells should be evaluated.
Well-characterized in vitro bioassays for LH have been developed.
Rat or mouse Leydig cells are incubated over several hours with
various concentrations of LH. The amount of testosterone produced
by the Leydig cells is measured. A potential testicular toxin could be
studied by exposing it, in various concentrations, with LH to the
-------
62
Leydig cell preparations and comparing the amount of testosterone
produced to that produced by cells exposed to LH alone.
10. The responsiveness of Sertoli cells should be evaluated.
Sertoli cell cultures, preferably from postpubertal males, can be used.
The production of known secretory products, such as transferrin or
androgen-binding protein, following stimulation by FSH, can be
measured as an index of their activity. An effect of a toxin would be
reflected in that index. A major question is the relationship of
in vitro toxicity to in vivo toxicity, particularly considering the
short-term nature of the tests and the long-term nature of in vivo
exposure. However, some of the in vitro tests provide limited
opportunity to evaluate human tissue directly with animal models.
11. Competitive mating (heterospermic insemination) should be
evaluated as a screening assay. The use of a mixed-insemination assay
(91—93) for screening toxicants offers a means of increasing the
sensitivity of fertilization assays and should be explored. Rabbits and
possibly rats could be used. Semen from exposed and control males
would be mixed and inseminated into the same female and the
paternity of the offspring established by genetic markers (i.e., eye or
coat color). This has good potential for use as a screening assay of
superior sensitivity which could simultaneously assess disturbances of
sexual behavior, sperm quality, sperm transport in the female,
fertilization, and embryonic and fetal development. A limited
number of trials with the system should be adequate to determine its
utility.
12. The direct assessment of damage to the sperm genome would
permit routine screening and monitoring of males for exposure in the
workplace to chemicals that may be hazards to their reproductive
capacity. Further studies might attempt the following:
(a) Establish the degree of correlation between abnormal sperm
head morphology and an aberrant chromosome complement.
(b) Develop sensitive methods for the identification and measure-
ment of alkylated or modified DNA bases.
(c) Improve the methodology to quantitate alkylated amino
acids, since there is evidence that alkylation of sperm chromatin
proteins also contributes to reproduction failure.
(d) Develop methods to detect damage to sperm chromatin (e.g.,
enzymatic detection of strand breaks in sperm DNA).
Additional studies could be designed and sponsored to evaluate
the suitability of the four techniques discussed above as routine
procedures for detection of genetic abnormalities by direct observa-
tion of spermatozoa and of the male pronucleus.
-------
63
REFERENCES
1. Smith, K. D., Rodriguez-Rigau, L. J. and Steinberger, E.: Relation between
indices of semen analysis and pregnancy rate in infertile couples. Fertil.
Steril.28: 1314-1319,1977.
2. Amann, R. P.: A critical review of methods for evaluation of spermato-
genesis from seminal characteristics. J. Androl. 2: 39—60, 1981.
3. Barfield, A., Melo, J., Coutinho, E., Alvarezs, P., Faundes, A., Brache, V.,
Leon, P., Frick, J., Bartsch, G. and Weiske,W. H.: Pregnancies associated
with sperm concentrations below 10 million-mi in clinical studies of a
potential male contraceptive method, monthly depot medroxyproges-
terone acetate and testosterone esters. Contraception 20(2): 121—127,
1979.
4. David, G., Jouannet, P., Martin-Boyee, A., Spira, A. and Schwartz, D.:
Sperm counts in fertile and infertile men. Fertil. Steril. 31(4): 453—455,
1979.
S.Zuckerman, Z., Rodriguez-Rigau, L. J., Weiss, D. B., Chowdhury, A. K.,
Smith, K. D. and Steinberger, E.: Quantitative analysis of the seminiferous
epithelium in- human testicular biopsies and the relation of spermato-
genesis to sperm density. Fertil. Steril. 30: 448-455, 1978.
6. Freund, M.: Standards for the rating of human sperm morphology. Int. J.
Fertil. 11:97-180,1966.
7. Johnson, L., Petty, C. S. and Neaves, W. B.: A comparative study of daily
sperm production and testicular composition in humans and rats. Biol.
Reprod. 22: 1233-1243,1980.
8. Amann, R. P. and Howards, S. S.: Daily spermatozoal production and
epididymal spermatozoal reserves of the human male. J. Urol. 124:
211-215,1980.
9. Smith, K. D. and Steinberger, E.: What is oligospermia? In: The Testis in
Normal and Infertile Men, D. Troen and H. Nankin, Eds., Raven Press:
New York; pp. 489-503, 1977.
10. Salisbury, G. W., Van Demark, N. L. and Lodge, L. R.: Physiology of
reproduction and artificial insemination of cattle, W. H. Freeman and Co.:
San Francisco; 788 pp., 1978.
11. MacLeod, J., Hotchkiss, R. S. and Sitterson, B. W.: Recovery of male
fertility after sterilization by nuclear radiation. J. Am. Med. Assoc. 187:
637-641,1964.
12. Amelar, R. D., Dubin, L. and Hotchkiss, R. S.: Restoration of fertility
following unilateral orchiectomy and radiation therapy for testicular
tumors. J. Urol. 106: 714-718, 1971.
13. Meistrich, M. L., Hunter, N., Suzuki, N., Trostle, P. K. and Withers, H. R.:
Gradual regeneration of mouse testicular stem cells after ionizing
radiation. Radiat. Res. 74: 349-362, 1978.
14. Rowley, M. J., Leach, D. R., Warner, G. A. and Heller, C. G.: Effect of
graded doses of ionizing radiation on human testes. Radiat. Res. 59:
665-678, 1974.
-------
64
IS.Cheviakoff, S., Calamera, J. C., Morgenfield, M. and Mancini, R. E.:
Spermatogenesis in patients with lymphoma after treatment with chlo-
rambucil.J.Reprod. Fertil. 33: 155-157,1973.
16. Sherins, R. J. and DeVita, V. T., Jr.: Effect of drug treatment for lymphoma
on male reproductive capacity—studies of men in remission after therapy.
Ann. Intern.Med. 79: 216-220,1973.
17. Sieber, S. M. and Adamson, R. H.: Toxicity of antineoplastic agents in man:
chromosomal aberrations, antifertility effects, congenital malformations,
and carcinogenic potential. Adv. Cancer Res. 22: 57—155,1975.
18. Lu, C. C. and Meistrich, M. L.: Cytotoxic effects of chemotherapeutic drugs
on mouse testis cells. Cancer Res. 39: 3575-3582,1979.
19.Torkelson, T. R., Sadele, S. E., Rowe, V. K., Kodama, J. K., Anderson,
H.H., Loquvam, G.S. and Hine, C.H.: Toxicologic investigations of
l,2-dibromo-3-chloropropane. Toxicol. Appl. Pharmacol. 3: 545-559,
1961.
20. Whorton, M. D., Krauss, R. M., Marshall, S. and Milby, T. H.: Infertility in
male pesticide workers. Lancet 2(8051): 1259-1261,1977.
21. Glass, R. I., Lyness, R. N., Mengle, D. C., Powell, K. E. and Kahn, E.: Sperm
count depression in pesticide applicators exposed to dibromochloropro-
pane. Am. J.Epidemiol. 109(3): 346-351,1979.
22. Lantz, G. D., Cunningham, G. R., Huckins, C. and Lipshultz, L. I.: Recovery
from severe oligospermia after exposure to dibromochloropropane. Fertil.
Steril. 35: 46-53,1981.
23.Amann, R. P.: Use of animal models for detecting specific alterations in
reproduction. Appl. Fund. Toxicol.; in press, 1982.
24. Amann, R. P.: Sperm production rates. In: The Testis, Vol. 1, A. Johnson,
W. Gomes, and N. Van Demark, Eds., Academic Press, Inc.: New York;
pp. 433-482,1970.
25. Carson, W. S. and Amann, R. P.: The male rabbit. VI. Effects of ejaculation
and season on testicular size and function. J. Anim. Sci. 32: 302—309,
1972.
26. Foote, R. H.: Research techniques to study reproductive physiology in the
male. In: Techniques and Procedures in Animal Science Research.
American Society of Animal Producers: Albany, New York; pp. 81 — 100,
1969.
27. Robb, G. W., Amann, R. P. and Killian, G. J.: Daily sperm production and
epididymal sperm reserves of pubertal and adult rats. J. Reprod. Fertil. 54:
103-107,1978.
28. Amann, R. P. and Lambiase, J. T.: The male rabbit. III. Determination of
daily sperm production by means of testicular homogenates. J. Anim. Sci.
28: 369-374,1969.
29. Amann, R. P., Kavanaugh, J. F., Griel, L. C., Jr. and Voglmayr, J. K.: Sperm
production of Holstein bulls determined from testicular spermatid
reserves, after cannulation of rete testis or vas deferens, and by daily
ejaculation. J. Dairy Sci. 57: 93-99,1974.
-------
65
30t Johnson, L., Petty, C. S. and Neaves, W. B.: A new approach to
quantification of spermatogenesis and its application to germinal cell
attrition during human spermiogenesis. Biol. Reprod. 25: 217—226, 1981.
Sl.Clermont, Y.: Kinetics of spermatogenesis in mammals: seminiferous
epithelial cycles and spermatogonial renewal. Physiol. Rev. 52: 198—236,
1972.
32. Swierstra, E. E. and Foote, R. H.: Cytology and kinetics of spermatogenesis
in the rabbit. J. Reprod. Fertil. 5: 309-322, 1963.
33. Swierstra, E. E., Whitefield, J. W. and Foote, R. H.: Action of amphotericin
B (fungizone) on spermatogenesis in the rabbit. J. Reprod. Fertil. 7:
13-19,1964.
34. Berndtson, W. E.:- Methods for quantifying mammalian spermatogenesis: a
review. J. Anim. Sci. 44: 819-833, 1977.
35. Lu, C. C., Meistrich, M. L. and Thames, H. D.: Survival of mouse testicular
stem cells after gamma or neutron irradiation. Radiat. Res. 81: 402—415,
1980.
36. Amann, R. P.: The male rabbit. IV. Quantitative testicular histology and
comparisons between daily sperm production as determined histologically
and daily sperm output. Fertil. Steril. 21: 662-672,1970.
37. Lambiase, J. T., Jr. and Amann, R. P.: The male rabbit. V. Changes in sperm
reserves and resoiption rate induced by ejaculation and sexual rest. J.
Anim. Sci. 28: 542-549, 1969.
38. Fray, C. S., Hoffes, A. P. and Fawcett, D. W.: Reexamination of motility
patterns of rat epididymal spermatozoa. Anat. Rec. 173: 301—308, 1972.
39. Gaddum, P.: Sperm maturation in male reproductive tract—development of
motility. Anat. Rec. 161: 471-482, 1968.
40. Blandau, R. J. and Rumery, R. E.: Relationship of swimming movements of
epididymal spermatozoa to their fertilizing capacity. Fertil. Steril. 15:
571-579,1964.
41. Lock, L. F. and Scares, E. R.: Increases in morphologically abnormal sperm
in rats exposed to 60Co irradiation. Environ. Mutagen. 2: 125-131, 1980.
42. Harasymowyez, J., Ball, L. and Seidel, G. E., Jr.: Evaluation of bovine
spermatozoal morphologic features after staining or fixation. Am. J. Vet.
Res. 37: 1053-1057,1976.
43. Holtz, W. and Foote, R. H.: Anatomy of reproductive system in male Dutch
rabbits (Oryctolagus cuniculum) with special emphasis on the accessory
sex glands. J. Morph. 158: 1-20, 1978.
44. Amann, R. P.: Effect of ejaculation frequency and breed on semen
characteristics and sperm output of rabbits. J. Reprod. Fertil. 11:
291-293,1966.
45. Holtz, W. and Foote, R. H.: Sperm production, output, and urinary loss in
the rabbit. Proc. Soc. Exp. Biol. Med. 141: 958-962,1972.
46.Kirton, K. E., Desjardins, C. and Hafs, H. D.: Levels of some normal
constituents of rabbit semen during repetitive ejaculation. Fertil. Steril.
17:204-211,1966.
-------
66
47.Paufler, S. K. and Foote, R. H.: Effect of triethylenemelamine (TEM) and--
cadmium chloride on spermatogenesis in rabbits. J. Reprod. Fertil. 19:
309-319,1969.
48.Schwartz, D., Laplanche, A., Jouannet, P. and David, G.: Within-subject
variability of human semen in regard to sperm count, volume, total
number of spermatozoa, and length of abstinence. J. Reprod. Fertil. 57:
391-395,1979.
49. Katz, D. F. and Overstreet, J. W.: Sperm motility assessment by videomicro-
graphy. Fertil. Steril. 35(2): 188-193,1981.
50. Eliasson, R.: Analyses of semen. In: Progress in Infertility, S. J. Behram and
R. W. Kistnes, Eds., Little Brown and Co.: Boston; pp. 691-713, 1975.
51. Holtz, W. and Foote, R. H.: Composition of rabbit semen and the origin of
several constituents. Biol. Reprod. 18: 286-292, 1978.
52. Baker, M. W. G., Bremner, W. J., Burger, H. G., de Kretser, D. M., Dulmanis,
A., Eddie, L., Hudson, B., Keogh, E.J., Lee, V.W.K. and Rennie,G.:
Testicular control of FSH secretion. Rec. Prog. Horm. Res. 32: 429, 1976.
53. de Kretser, D. M., Kerr, J. B., Rick, K. A., Risbridger, G. and Dobas, M.:
Hormonal factor involved in normal spermatogenesis. In: Testicular
Development, Structure, and Function, A. Steinberger and E. Steinberger,
Eds., Raven Press: New York; pp. 109-115,1980.
54.Psychoyos, A.: Hormonal control of ovoimplantation. Vitam. Horm. 31:
201-256, 1973.
55.Bedford, J. M.: Techniques and criteria used in the study of mammalian
fertilization. In: Methods in Mammalian Embryology, J.C. David, Ed.,
W. H. Freeman and Co.: San Francisco; pp. 37—63, 1971.
56. Yanagimachi, R., Yanagimachi, H. and Rogers, B. J.: Use of zona-free animal
ova as a test system for assessment of fertilizing capacity of human
spermatozoa. Biol. Reprod. 15(4): 471-476, 1976.
57.Hahn, J., Foote, R. H. and Cranch, E. T.: Tonometer for measuring
testicular consistency of bulls to predict semen quality. J. Anim. Sci. 29:
483-489,1969.
58. Oakberg, E. F.: Degeneration of spermatogonia of the mouse following
exposure to X-rays, and stages in the mitotic cycle at which death occurs.
J.Morph. 97: 39-54,1955.
59.Nicander, L.: Studies on the regional histology and cytochemistry of the
ductus epididymis in rabbits. Acta Morphol. Neerl. Scand. 1: 99—118,
1957.
60. Reid, B. L. and Cleland, K. W.: The structure and function of the
epididymis. I. The histology of the rat epididymis. Aust. J. Zool. 5:
223-246,1957.
61. Brooks, D. E.: Biochemical environment of sperm maturation. In: The
Spermatozoon: Maturation, Motility, Surface Properties and Comparative
Aspects, D.W. Fawcett and J. M. Bedford, Eds., Urban and Schwarzen-
berg: Baltimore; pp. 23-34,1979.
62. Brandes, D., Ed.: Male Accessory Sex Organs. Academic Press: New York;
527 pp., 1974.
-------
67
63.Hammerstedt, R. H.: Characterization of sperm surfaces using physical
techniques. In: Spermatozoon: Maturation, Motility, and Surface Proper-
ties and Comparative Aspects, D.W. Fawcett and J.M. Bedford, Eds.,
Urban and Schwarzenberg: Baltimore; pp. 205—216,1979.
64. Bedford, J. M. and Cooper, G. W.: Membrane fusion events in the
fertilization of vertebrate egg. In: Cell Surface Reviews, Vol. 5, Membrane
Fusion, Elsevier North Holland: Amsterdam; pp. 66—111,1978.
65.Mann, T. and Lutwak-Mann, C.: Male Reproductive Function and Semen,
Springer-Verlag: New York; 495 pp., 1981.
66. Hammerstedt, R. H.: Monitoring the metabolic rate of germ cells and sperm.
In: Reproductive Processes and Contraception, K.W. McKerns, Ed.,
Plenum Publishing: New York; pp. 353-391, 1981.
67. Hegde, U. C., Shastry, P. R. and Rao, S. S.: Use of dithiothreitol for
improved visibility of F-body in human Y-bearing spermatozoa. J. Reprod.
Fertil. 53: 403-405, 1978.
68. Evenson, D. P., Darzynkiewicz, Z. and Melamed, M. R.: Relation of
mammalian sperm chromation heterogeneity to fertility. Science 210:
1131-1133,1980.
69.Gledhill, B. L., Lake, S. and Dean, P. N.: Flow cytometry and sorting of
sperm and other male germ cells. In: Flow Cytometry and Sorting,
M.Melamed, P.Mullaney, and M.Mendelsohn, Eds., J. Wiley and Sons:
New York; pp. 471-484,1979.
70. Rudak, E., Jacobs, P. A. and Yanagimachi, R.: Direct analysis of
chromosome constitution of human spermatozoa. Nature 274: 911—913,
1978.
71. Aafjes, J. H., Vels, J. M. and Schenck, E.: Fertility of rats with artificial
oligozoospermia. J. Reprod. Fertil. 58(2): 345-352, 1980.
72.Desjardins, C., Kirton, K. T. and Hafs, H. D.: Sperm output of rabbits at
various ejaculation frequencies and their use in the design of experiments.
J. Reprod. Fertil. 15: 27-32, 1968.
73. Seidel, G. E., Jr. and Foote, R. H.: Variance components of semen criteria
from bulls ejaculated frequently and their use in experimental design. J.
Dairy Sci. 56: 399-405,1973.
74. MacLeod, J.: Human male infertility. Obstet. Gynecol. Surv. 26: 335—351,
1971.
75. Janick, J. and MacLeod, J.: Measurement of human spermatozoan mobility.
Fertil. Steril. 21: 140-146,1970.
76. Lubs, H. A.: Testicular size in Kleinfelter's Syndrome in men over fifty.
Report of a case with XXY/XY mosaicism. New Engl. J. Med. 267:
326-331,1962.
77. Asbjornsen, G., Molne, K., Klepp, O. and Aakvaag, A.: Testicular function
after combination chemotherapy for Hodgkin's disease. Scand. J. Hae-
matol. 16(1): 66-69,1976.
78..Van Thiel, D. H., Sherins, R. J., Meyers, G. H. and Devita^ V.T., Jr.:
Evidence for a specific seminiferous tubular factor affecting follicle-
stimulating hormone secretion in man. J. Clin. Invest. 51: 1009—1019,
1972.
-------
68
79. Cunningham, G. R. and Huckins, C.: Serum FSH, LH, and testosterone in
60Co 7-irradiated male rats. Radiat. Res. 76: 331-338,1978.
80. Verjans, H. L. and Eik-Nes, K. B.: Hypothalamic-pituitary-testicular system
following testicular X-irradiation. Acta Endorcinol. 83(1): 190-200,
1976.
81. MacLeod, J.: Human seminal cytology following the administration of
certain antispermatogenic compounds. In: Agents Affecting Fertility,
C. R. Austin and J. S. Perry, Eds., Little Brown and Co.: Boston; pp.
93-123,1965.
82. Belsey, M. A., Eliasson, R., Gallegos, A. J., Maghissi, K. S., Paulsen, C. A.
and Prasad, M. R.N.: Laboratory Manual for the Examination of Human
Semen and Semen-Cervical Mucus Interaction, Press Concern: Singapore;
43 pp., 1980.
83. National Academy of Sciences: Drinking Water and Health, Vol. 3, National
Academy Press: Washington, D.C.; pp. 25-265,1980.
84. Young, W. C., Ed.: Sex and Internal Secretions, Vol. 1,3rd Edition, Williams
andWilkins: Baltimore;pp. 161-448,1961.
85. Young, W. C., Ed.: Sex andlnternal Secretions,Vol. 2,3rd Edition, Williams
andWilkins: Baltimore;pp.707-796,1173-1239,1961.
86. Mann, T.: The Biochemistry of Semen and of the Male Reproductive Tract,
Methuen: London; 240 pp., 1964.
87. Hamilton, D. W. and Creep, R. O., Eds.: Handbook of Physiology, Section 7,
Vol. 5: Male Reproductive System, American Physiological Society:
Washington,D.C.; 519 pp., 1975.
88. Creep, R. 0., Koblinsky, M. A. and Jaffee, F. S., Eds.: Reproduction and
Human Welfare: A Challenge to Research, MIT Press: Cambridge, Mass.;
622 pp., 1976.
89.Wyrobek, A. J. and Gledhill, B. L.: Human semen assays for workplace
monitoring. In: Proceedings' of Workshop on Methodology for Assessing
Reproductive Hazards in the Workplace. Center for Disease Control,
National Institute for Occupational Health and Safety; pp. 327—355,
1980.
90. Gledhill, B. L. Personal communication, Lawrence Livermore Laboratory:
California; 1981.
91. Beatty, R. A., Bennett, G. H., Hall, J. G., Hancock, J. L. and Stewart, D. L.:
An experiment with heterospermic insemination in cattle. J. Reprod.
Fertil. 19: 491-502,1969.
92. Overstreet, J. W. and Adams, C. E.: Mechanisms of selective fertilization in
the rabbit: sperm transport and viability. J. Reprod. Fertil. 26: 219—231,
1971.
93. O'Connor, M. T., Amann, R. P. and Saacke, R. G.: Comparisons of computer
evaluations of spermatozoa! motility with standard laboratory tests and
their use for predicting fertility. J. Animal Sci.; in press, 1981.
-------
APPENDIX
DETAILS OF TEST PROTOCOLS AND
GLOSSARY OF TERMS FOR MALE
RISK ASSESSMENT
I. DESCRIPTION AND DISCUSSION OF TESTS USEFUL
IN ANIMAL MODELS OR MAN
Body Weight
Measure body weight of all test animals weekly starting two
weeks before administration of the compound and continuing until
termination of the study.
Testicular Characteristics
Testis size in situ
The number of spermatozoa, and to a lesser extent the quantity
of testosterone, produced by the testes of normal individuals is a
function of testis size and, to a lesser extent, of variation in the
proportion of the testis composed of germinal elements and
interstitial tissue (1). Therefore, assessment of testicular size is very
important from a functional standpoint. In scrota! animals, testicular
size can be measured easily, accurately, repeatedly and without
damage to the individual (2—5). In many species of laboratory and
domestic animals, testis size is correlated (correlation coefficient
r = 0.8—0.9) with sperm output in ejaculated semen when males are
ejaculated frequently (e.g., four ejaculates per week) (1, 2). Changes
in testis size should be correlated with results of other tests to
increase the accuracy of the analysis.
Measurements of testis size should be made biweekly or weekly
with animal models and could be made part of an annual physical
examination given to men working in a hazardous environment.
69
-------
70
Scrotal circumference. This measurement can be taken easily in
animals with pendulous scrotum (4). To reduce variation, the
measurements should be made with a standard procedure.
Linear measurements. Length and width can be measured in
species such as dogs, rabbits, bulls, and horses (1, 2—5). Length and
width measurements are correlated with testis weight O0.90). If
these data are correlated with seminal characteristics, adjustment for
time lag in spermatogenesis and sperm transport through the
excurrent ducts is necessary.
Testis weight
Each testis must be dissected free from the epididymis and
pampiniform plexus, and weighed when model animals are killed or
castrated. Testis weight, relative to norms for that breed or strain,
can reveal gross differences resulting from a treatment.
Spermatid reserves
Counting of homogenization-resistant spermatid nuclei in testicu-
lar homogenates is a simple, accurate, and sensitive method for
measuring sperm production. This method can be accomplished with
simple equipment and does not require extensive training. The nuclei
of elongated spermatids are resistant to mechanical and chemical
disruption and are easily identified after physical disruption of
testicular tissue (1,6, 7). With human testes, small biopsies can be
used (8), and the tissue should be fixed in glutaraldehyde before
homogenization, because some spermatid nuclei may not be fully
condensed (9). The interval from when spermatids acquire the
resistance to homogenization until spermiation is a constant for a
species or strain (1). Thus, the number of resistant spermatids is a
direct measure of the production of spermatozoa by the testis and
the survival of the precursor spermatogenic cells (1).
Either biopsy material (20 mg or more), a representative sample
taken at necropsy, or the entire testes (for rats and rabbits) can be
homogenized or disrupted ultrasonically (6, 10). Resistant spermatid
nuclei are counted in a cytometer (at least 6 chambers per sample).
Counts should be expressed on both a per-testis basis and a
per-mg-of-parenchyma basis (11). Counts from treated animals
should be compared to those for concurrent control males.
The time of this analysis relative to an acute treatment or the
onset of chronic treatment can be varied so as to reveal possible
damage to cells in specific stages of spermatogenesis. The interval
-------
71
between onset of treatment and evaluation should be expressed in
terms of the duration of one "cycle of the seminiferous epithelium"
for that species. Preferred times for evaluating agents should be
chosen according to the kinetics of spermatogenesis (7, 12—17).
Histopathological analysis of testes
Histologic analyses of testicular biopsies or whole testes must be
performed on animal models and, in special cases, could be
performed on man. Qualitative and quantitative analyses of in-
creasing complexity yield general or precise information. It is
axiomatic that serious disturbance will be evident to the observer
using direct simple evaluations of germinal epithelium in histological
preparations of whole testes. Threshold effects require a detailed
evaluation such as that recommended below. Electron microscopy is
not considered to be useful for screening of damaging agents.
Testicular tissue must be fixed immediately in Bouin's or Zenker's
fluid (10% formalin is not satisfactory). Slides should be stained with
hematoxylin and eosin for simple analyses and with periodic-acid-
Schiff-hematoxylin if a more precise determination of the stages of
spermatids is required (17).
Gross morphology. Appearance of Leydig interstitial cells (18);
occurrence of lymphoid cell or macrophage infiltration (19);
presence of germ cells of each stage (spermatogonia, spermatocytes,
spermatids, sperm) in seminiferous tubules (20); presence of large
numbers of degenerating (21), multinucleate (22), or abnormal germ
cells (23) should be noted.
Nonfunctional tubules. The percentage of tubular cross-sections
with no evidence of spermatogenesis (i.e., <4 germ cells) should be
scored during brief examination at 100X or 400X magnification of
250 cross-sections per testis (24, 25). Such examination could be
performed 2 to 7 days after acute treatment or 6 cycles after onset
of chronic treatment. The integrity of the layer of Sertoli cells in
these sterile tubules should also be noted.
Tubules with spermatids lining the lumen. The end product of
spermatogenesis is reflected in the "mature" spermatids about to be
released from the Sertoli cells. Tubules with spermatids aligned at the
lumen can be easily recognized (16). The incidence of such tubules is
a characteristic of the species. Deviations between control and
treated males reflect testicular dysfunction.
Seminiferous tubule diameter. Diameters can change with inter-
ference of tubular function (26). Measurements of minor diameter
should be taken on essentially round tubule cross-sections (cut at
-------
72
right angles to their long axis) taken from several different locations
in sections used for scanning other aspects of gross histology. Only
tubules in which the minor diameter is within 10% of the major
diameter (i.e., the sections are essentially transverse) should be
measured.
Counts of preleptotene or leptotene spermatocytes
The number of leptotene spermatocytes per Sertoli cell with a
visible nucleolus can be quantitated because of a characteristic
nuclear morphology of leptotene spermatocytes (27, 28). The
number of leptotene spermatocytes and the number of Sertoli cells
with a visible nucleolus should be determined in the same set of
tubules. The ratio of spermatocytes per Sertoli cell is a sensitive
measure of testicular damage; effects of as little as 5 rad of radiation
can be detected (20).
Epididymal Characteristics
Weight of distal half of epididymis
The distal portion of the epididymis can be isolated by severing
the corpus epididymidis midway between the caput and cauda and at
the junction of the distal cauda with the ductus deferens. The distal
epididymis and the contralateral epididymis should be weighed
promptly.
Number of sperm in the distal half of epididymis
One epididymis, weighed as above, is homogenized to liberate the
spermatozoa contained therein (6, 7); simple mincing of the tissue is
inadequate. Sperm cells are counted using a cytometer (at least 6
chambers counted per sample). The results should be expressed as
total counts. The epididymis evaluated could be alternated within
each control or treated group to ensure representative sampling if
there is any systematic difference between sides.
Motility of sperm from the distal end
Sperm from the distal end of the remaining cauda epididymidis
will be expressed into a phosphate-buffered saline solution contain-
ing 5 mM of glucose or pyruvate plus 0.1% bovine serum albumin,
polyvinyl alcohol, or similar macromolecules. Sperm concentration
should be standardized to 10 X 106 to 40 X 106 per ml, and the
-------
73
percentage of motile sperm determined at 37°C under conditions
similar to those described for estimating percentage of motile
spermatozoa in ejaculated semen.
Gross morphology of spermatozoa from the distal end
The same semen preparation used for estimating the percentage
of motile spermatozoa will be viewed by phase-contrast microscopy
at 400X for evaluation of gross morphology. The proportion of the
spermatozoa from treated animals, in comparison with controls, with
misshapen heads, acrosomal defects or distorted swimming patterns,
will be estimated.
Detailed morphology of spermatozoa from the distal end
For detailed evaluation of sperm morphology, smears will be
prepared by a procedure minimizing artifacts (29), then stained with
eosin-nigrosin (or other differential stain). A total of 200 to 400
spermatozoa should be classified per sample. Smears can be preserved
as a permanent record, pr videotapes can be prepared. Detailed
morphological or morphometric examination is possible with either.
Accessory Sex Gland Characteristics
1. The accessory sex glands are biomonitors of androgen produc-
tion by the testes. Thus, accessory sex gland weight will be recorded
when each male animal is killed.
2. For rats, the vesicular glands are discrete organs and very easily
distinguishable. After removal and expression of the viscid fluid, the
glands should be blotted and weighed.
3. The individual accessory sex glands are not discrete in the
rabbit (30). Thus, the total set of accessory glands will be excised as
a single unit, blotted, and weighed. The organs will be reweighed
after removal of any secretion present in the vesicular glands.
Seminal Analysis
General aspects of seminal analysis
(a) Analysis of semen offers a convenient approach for monitor-
ing function of the germinal epithelium and, with less specificity, the
functions of the epididymides, prostate, vesicular glands, and
bulbourethral glands (11). An abnormality in epididymal function
may be detected in semen ejaculated 3 to 15 days after epididymal
-------
74
dysfunction. An abnormality in spermatogenesis typically cannot be
detected in semen until after at least 1 to 4 cycles of the
seminiferous epithelium (16-64 days in man or 11-43 days in
rabbit) have passed (1, 11, 14,31), plus time for epididymal trans-
port. This long interval is required because an agent must accumulate
to a toxic concentration and produce a lesion in germ cells at a
specific point in their development (often >2 cycles of the
seminiferous epithelium before the end of spermatogenesis) before
production of more mature germ cells is affected. After passage of a
given interval, evidence of the lesion can be seen within the testis,
but the lesion will not be evident in semen until the affected germ
cells have completed spermatogenesis (2—3 cycles), passed through
the epididymis (4—16 days), and appeared as spermatozoa in
ejaculated semen.
(b) Multiple samples can and should be obtained from each
individual male. Both quantitative and qualitative characteristics of
more than one ejaculate must be evaluated to gain a reasonable
understanding of testicular function (1, 11,32). Data for samples
collected before experimental exposure can be used as one basis for
assigning males to control or treatment groups or as a covariant in
the statistical analysis.
(c) The species, strain, and age of males; testicular size; season;
method of semen collection; and interval since the previous
ejaculation(s) all influence quantitative characteristics of semen and
must be carefully controlled (11, 33).
(d) If seminal analyses are planned, use of a species from which
semen can be collected by artificial vagina or digital manipulation
(masturbation) is essential. Suitable species include man, rabbit, dog,
bull, and minipig (1,2, 11,32-34). The rabbit is the species of
choice for screening potentially toxic agents because of size,
availability, cost, and ease of use. Small rabbits (e.g., Dutch Belted)
are as good as larger breeds (e.g., New Zealand White) and are
cheaper to house. Rams, goats, and stallions are less ideal because
seasonal changes are more profound. Subhuman primates probably
will not be used frequently because of their limited numbers and
cost. Although useful in many other aspects, rats have limited use (as
do mice) because of their small seminal volume, difficulty in
quantifying seminal characteristics, and the necessity to use electro-
stimulation for semen collection. Improved procedures for quantita-
tive collection of semen from rats or mice are unlikely to overcome
their limitations. However, as outlined above, cauda epididymal
sperm can be obtained (on a one-time basis) from a mouse or rat and
-------
75
evaluation of the motility and morphology of epididymal sperm is
desirable.
(e) For most studies, sexually mature males (body weight <90%
of maximum value for that strain) should be used. If consequences of
exposure before puberty are to be evaluated, the age or body weight
when a given number of sperm are first ejaculated may be a useful
criterion (requires at least weekly testing) and the postpubertal
changes in semen quality could be monitored (33).
By monitoring seminal characteristics longitudinally from ex-
posure, through a reasonable interval when an effect might be
expressed (6 X the duration of the cycle of the seminiferous
epithelium for the species studied) and during a recovery phase (if
appropriate) of twice this duration, information can be obtained on
the point when damage is expressed and when recovery occurs
(1, 2, 14).
Volume
(a) Volume of the ejaculate should be measured with an accuracy
of greater than 90 to 95%. To measure accurately ejaculates with a
small volume, it is recommended that collection tubes be preweighed
and ejaculate volume be calculated from the weight, assuming a
specific gravity of 1.0.
(b) Systematic errors associated with seminal loss during collec-
tion or transfer to a measuring device should be minimized.
Measurement of ejaculate volume within the collection vessel, after
addition of a known volume of buffer if essential, is desirable.
Systematic errors in measurement often can be corrected for
(1,2,11).
(c) If a uniform collection interval and standardized collection
procedure are used (1, 11), differences of >25% in ejaculate volume
probably could be detected in a longitudinal study utilizing ten
rabbits per treatment group (35). A difference of this magnitude
probably would reflect abnormal function of the accessory sex
glands if unaccompanied by a change hi sperm output.
(d) The coefficient of variation for volume of a human ejaculate
is unknown but could be calculated from available data. It is likely
that a sizable number of ejaculates must be evaluated for each
individual in a group to detect a 25% change in seminal volume.
Seminal plasma constituents
(a) The biochemical components of seminal plasma may reflect
the functionality of the epididymides and accessory sex glands (11),
-------
76
but the concentration of a compound in a seminal plasma has limited
diagnostic value (11).
(b) If collection procedures are rigidly standardized for a species,
a marked change (>25%) in the total mass of a constituent
ejaculated,-could reflect the function of the excurrent duct system or
one or more accessory sex glands. It is not clear at present, however,
whether any constituents change their concentration markedly in the
course of repeated ejaculations with humans (see Research Needed).
Spermatozoal concentration
(a) The term spermatozoal concentration is preferred to those of
sperm count or sperm density.
(b) Sperm concentration, by itself, provides little information
(11), but sperm concentration must be determined accurately so that
the total number of sperm per ejaculate can be calculated (see
below).
(c) Sperm concentration should be determined using a calibrated
spectrophotometer or electronic cell counter, if contaminating cells
or debris are not a problem, because of their accuracy and precision
(2, 11). If extraneous material is present in the semen, visual counts
using a cytometer are essential. Use of a cytometer (with a
phase-contrast microscope) is time consuming, and >6 replicate
counts are necessary to achieve >90% accuracy for a single sample.
Total sperm per ejaculate
(a) The term total number of sperm per ejaculate (vol-
ume X sperm concentration) is preferable to that of total sperm
count.
(b) Total sperm per ejaculate represents the number of sperm
coming, from the excurrent duct system and is independent of the
degree of dilution by accessory sex gland fluid (11).
(c) When semen is collected by a uniform procedure and total
sperm per ejaculate is averaged over time, daily sperm output
(number of sperm in a series of ejaculates divided by the time span)
can be calculated. Daily sperm output, in rabbits and bulls, is highly
correlated (sO.9) with daily sperm production (1, 11, 36).
(d) To measure daily sperm output accurately (1, 11), a uniform
interval of one, two, or three days between semen collections is
essential, and the series of ejaculates should extend over 14
(preferably 20) days (data for the first 3—6 ejaculates should be
excluded a priori and data for the remaining >6 samples averaged).
-------
77
(e) If semen is collected infrequently (one ejaculate weekly), a
50% reduction in sperm production probably would be undetectable
(11). To have a 75% chance of detecting a difference of 50% daily
sperm output would require about 20 rabbits per treatment and
ejaculation for >5 weeks (35).
(f) A formula relating the coefficient of variation (CV) between
counts; the significance level at which the statistical test is to be
performed, a (e.g., a = 0.05, a =0.01); the desired power or
sensitivity of the statistical test, 1-/3 (e.g., 1-/3 = 0.50, l-j3 = 0.90); and
the sample size of the study, N (i.e., N treated and N control
animals), to the required change (in terms of percent of 'normal' or
control values) in the test criteria is given by
= % change
VN/2
where Za = 1.645 for a = 0.05 and Za = 2.326 for a = 0.01, and
Zj.0 = 0, 0.253, 0.524, 0.842, and 1.281 for 1-0 = 0.5, 0.6, 0.7, 0.8,
and 0.9, respectively. It should be noted that this formula assumes a
one-sided statistical test, that is, looking for changes between treated
and control animals in only one direction (e.g., decrease in sperm
concentration). Along with the coefficients of variation given in
Table 10, it can be used to determine the adequacy of different
experimental designs. For example, the largest coefficient of varia-
tion, other than for accessory sex gland weight, is for the test
criterion total sperm/ejaculate in Dutch Belted rabbits, CV = 0.75.
Assuming that any statistical comparison between 12 treated and 12
control rabbits is 'conducted at the a = 0.05 level (i.e., 5% test level),
to have at least a 50% chance of detecting a statistically significant
difference (i.e., power = 0.50), then the treatment must produce at
least a 50% change in the test criterion, that is,
(1.654 + 0)(0.75) V 12/2 = 0.5. Because each of the other criteria,
except for accessory sex glands, have coefficients of variation of less
than 0.75, they would have the same power, 50%, of detecting a
smaller effect; for example, for testis weight, CV = 0.2 giving a
percent change of 13%, (1.645 + 0) (0.20)A/12/2 = 0.13.
Sperm motility
(a) Rigid control of temperature at 37°C and other conditions of
evaluation are essential (2, 11).
(b) Visual evaluation of sperm motility using diluted semen and a
phase-contrast microscope is informative and rapid, although sub-
-------
78
jective. Visual estimations are adequate for an initial screen, provided
that control and treatment samples are presented randomly in a blind
manner to the observer. For second- or third-level analysis, more
objective procedures, such as videotape and analyses or track
motility (37—39) should be considered.
(c) The percentage of progressively motile sperm, the translatory
velocity, and the presence of sperm moving in a circular pattern or
backward should be recorded.
(d) A reduced percentage of motile sperm might reflect abnormal
spermatogenesis, abnormal functions of the epididymis or entrance
of an antimotility factor into the semen (via the excurrent ducts,
prostate, bulbourethral glands, or vesicular glands) where it could
exert a direct effect on the sperm.
(e)The percentage of motile sperm probably reflects both
normality of spermatogenesis and sperm metabolism.
(f) Variation within males in the percentage of motile sperma-
tozoa (and probably velocity) is less than for ejaculate volume or
total sperm per ejaculate (35, 40).
(g) A significant decrease in the percentage of motile sperm
would be a strong indicator for a potential decline in fertility and
especially so if sperm numbers are limited.
Spermatozoal morphology
(a) Abnormalities of sperm morphology reflect dysfunction of
the germinal epithelium (primary abnormality) or of the excurrent
duct system (secondary abnormality). Certain abnormalities cannot
be clearly attributed to a specific site of action.
(b) An increase in the percentage of abnormal sperm may
precede a decline in the total number of sperm per ejaculate (if any)
and can serve as a sensitive indicator of epididymal or testicular
function.
(c) Within a male, sperm morphology is quite consistent over
time (41). This consistency makes sperm morphology a sensitive
probe while requiring fewer samples per male for an experiment of a
given precision.
(d) Evaluation of sperm morphology is subjective (42) and must
be carefully standardized among laboratories (11, 42, 43). A detailed
classification probably is unnecessary in an initial screening process.
Classification of spermatozoa, based on light microscopy, as normal
or abnormal head, normal or abnormal tail is adequate for a screen.
For second- or third-level screening, a more complex classification
might be used (42-44).
-------
79
(e) Evaluation of sperm morphology using wet preparations and
phase-contrast microscopy is recommended for simplicity and
freedom from artifacts (11, 29), although preparation and retention
of stained smears (a simple stain like eosion-nigrosin or eosin-analine
blue) is desirable for archival purposes.
Ejaculated sperm as an in vitro test system
(a) Substances can pass from blood into semen via the fluid from
the. excurrent ducts and accessory sex glands and could be
spermicidal or alter sperm function.
(b) Agents can be screened economically by incubating sperm in
vitro under standard conditions in a protein-containing buffer at
37°C for 4 to 8 hours. Sperm could be exposed to the agent briefly
(10—30 minutes) or throughout the incubation period. A dose-
response curve should be established using objective methods of
evaluation and sperm from humans or other species (rats, rabbits,
dogs, or bulls).
(c) The decline in percentage of motile sperm over time is an
exceUent criterion, but other criteria (e.g., integrity of the acrosome
and plasma membrane, oxygen consumption, adenosine 5'-
triphosphate content, or degree of agglutination [45]) could be used.
(d) Compounds that are spermicidal in vitro at concentrations
that could be anticipated or shown to be present in blood or seminal
plasma should be carefully screened in vivo. Failure to demonstrate a
spermicidal action in vitro is not evidence that an agent would be
free of effects on male reproduction, nor is spermicidal action in
vitro evidence of in vivo activity.
Assessment of Male Reproductive Toxicity
Using Endocrinological Methods
General
(a) This section describes general aspects of applying endocrino-
logical methods to the study of male reproduction. Specific
applications of these techniques to studies in animals and men are
described elsewhere in this account.
Normal male reproductive function requires hormonal stimula-
tion of the testes and production by the testes of adequate numbers
of sperm and the hormone testosterone. Luteinizing hormone (LH)
and follicle-stimulating hormone (FSH) are the two hormones
necessary to maintain normal testicular function. These hormones
-------
80
originate in the pituitary gland and travel through the blood to affect
the testis. The production of LH and FSH and their release from the
pituitary are stimulated by gonadotropin-releasing hormone (GnRH),
which is produced in the hypothalamus at the base of the brain.
Testosterone and other hormones produced by the testis are carried
by the blood throughout the body. At the pituitary, these hormones
tend to decrease the production of LH and FSH, that is, they exert a
"negative feedback" effect on LH and FSH secretion.
(b) If a defect occurs in hypothalamic or pituitary function,
blood levels of FSH and LH will tend to decrease.
(c) If a defect occurs in the testis (either in sperm or testosterone
production), FSH and LH levels will tend to increase because of lack
of the "negative feedback" effect of testicular hormones.
(d) In addition to its effects on the pituitary, testosterone exerts
many effects throughout the body. It is necessary for expression of
male sexual behavior and the ability to perform intercourse,
stimulates muscle and bone development and red blood cell
production, and is essential for many other aspects of normal body
function. A decrease in blood levels of testosterone can be expected
to affect all these functions adversely.
(e) It is clearly established in all mammalian species investigated
that an endocrine defect in the brain, pituitary, or testis may inhibit
spermatogenesis and normal sexual behavior and cause sterility. Less
severe defects in these tissues (not so severe as to lead to infertility)
might be detected by measurements of hormone concentration in the
blood. In certain situations, including studies of human beings,
hormone measurements are very practical, because they can be
performed on ordinary samples of blood serum, whereas seminal
fluid or testicular tissue may be difficult or impossible to obtain.
(f) Hormonal measurements are important and sensitive tools in
the assessment of toxicity to the male reproductive system. They can
be compared directly among a variety of species and between control
and treated groups of any species including man. Hormonal data may
give a clue as to the tissue in which a toxic effect is occurring.
Hormone assay and application
(a) Hormones commonly are measured by radioimmunoassay.
This is an extremely sensitive technique and, when done in a
competent laboratory, is quite reliable for measuring testosterone,
FSH, and LH. Many commercial laboratories perform radio-
immunoassays for LH, FSH, and testosterone in human blood
samples.
-------
81
Developing these assays requires familiarity with the technique,
access to counters for radioactivity and availability of specialized
assay reagents such as purified preparations of the hormones and
antibodies to these hormones.
(b) Testosterone is an identical molecule in all species, so it can
be measured from any species in a single assay. Nevertheless,
appropriate species controls are essential to preclude the possibility
that a cross-reacting molecule is altering test results. Both LH and
FSH are protein hormones that differ slightly in molecular structure
among species. Assays for LH and for FSH tend to be species-
specific, so that assay reagents appropriate for the species in question
generally are required for measuring LH or FSH. For rats and man
these reagents are widely available, and such reagents have recently
been introduced for rabbits. For normal adult male rats, however,
the concentration of LH in peripheral blood is usually below the
sensitivity level of available radioimmunoassays.
Measurement of blood levels of hormones yields a direct
assessment of the level of exposure necessary to produce a toxic
effect. The methodology is sufficiently accurate to detect changes of
20% in mean hormone levels using generally available numbers of
animals (e.g., 20 per group). Statistical adequacy for these assays
when performed competently is very good; hi general, within-assay
coefficients of variations are below 10%, and between-assay coeffi-
cients of variations are less than 15% (46). These assays are specific
measures of reproductive toxicity; disease of other organ systems will
not affect these measurements unless the disease concomitantly
affects the reproductive system.
Any statistically significant difference between hormone levels in
a comparison of control and exposed animals or men can be accepted
as strong evidence for a toxic effect of the exposure on reproduction.
If such an effect were established by animal studies, this could be
used as strong evidence that a similar effect would occur in human
beings. In essentially every instance studied in detail, agents found to
be testicular toxins in one species have a similar effect in other
species (47).
(c) Hormone levels may be measured in single blood samples,
although evaluation of several samples taken at two-hour intervals is
better. The time of day should be standardized because of diurnal
rhythms. Ordinarily, no special preparation is necessary concerning
diet or physical activity.
(d) For all known reproductive toxins, the damage to reproduc-
tion is reversible if the exposure level is so low as to produce only a
-------
82
minimally detectable effect. With a more severe insult or a prolonged
toxin exposure, the damage may be irreversible.
Examination of Known Toxic Exposures
Humans
Seminal fluid analyses. At least one and preferably five seminal
fluids obtained by masturbation at two-day intervals should be
submitted by each man. Precautions must be taken to ensure an
accurate measurement of seminal volume (11). The age of the
individuals and the abstinence interval between samples should also
be considered in the evaluation (48-50). Variability among ejacu-
lates from the same individual is influenced by the length of
abstinence (50). The semen should be analyzed for volume, sperm
concentration, and total sperm per ejaculate. Total sperm per
ejaculate must be calculated and compared with norms. Sperm
motility (at 37°C) should be objectively assessed using phase-contrast
microscopy; motility should be characterized in terms of percentage
of motility and velocity. If feasible, videotapes should be made of
the living sperm cells, with the tapes being subsequently analyzed in
a laboratory familiar with this technique. Seminal smears should be
fixed (29) for subsequent analysis. Tests of sperm function such as
penetration of zona-free hamster eggs may not be feasible for field
work. However, if persistent infertility remains undiagnosed after
completion of the other studies proposed, the hamster egg in vitro
penetration test (51) should be considered.
Blood hormone levels. Peripheral venous blood samples should be
obtained at a standardized time of day (preferably 0700-0900
hours) for measurement of serum luteinizing hormone (LH), follicle-
stimulating hormone (FSH), and testosterone by radioimmunoassay.
Hormone measurements on both the exposed and control groups
should be performed in the same laboratory. This laboratory must be
one that is recognized for reliability and that maintains careful
quality control records.
Gonadotropin-releasing hormone (GnRH) test. This test has
been demonstrated to be capable of detecting mild degrees of
primary testicular dysfunction insufficient to elevate basal hormone
levels out of the normal range (52). It is not necessary if the basal,
unstimulated-hormone levels are abnormal. Blood samples for mea-
surement of LH and FSH are obtained before and at 30, 60, and 90
minutes after administration of GnRH (100 Mg i.v:). Synthetic GnRH
is now available for use as an investigational new drug through several
-------
83
pharmaceutical companies and will probably be approved by the
Food and Drug Administration for general use.
Animal models
Rats. Twenty mature male rats of a highly fertile strain (body
weight must be >90% of adult normal for that strain) will be given
>0.5 of the maximum tolerated dose (MTD) by inhalation, intra-
peritoneal (i.p.) injection, drinking water, or gavage for a period of
exactly 6 cycles of seminiferous epithelium (12.9 days per
cycle X 6 = 77 days). An appropriate control group (or groups) will
be evaluated concurrently. If the agent is given by injection or
gavage, both nonhandled and vehicle-injected control groups are
necessary. Each rat will be weighed weekly starting 14 days prior to
initial dosing (day —14) and continuing to day 78. Each male will be
caged with two sexually mature, virgin female rats between days 65
and 71 (for a total of 6 nights) to evaluate fertility. On day 78, blood
will be taken by cardiac puncture immediately (<1 minute) after
removing a male rat from his cage in the animal room and the male
rat killed. Serum will be frozen.
Both testes will be weighed. One will be fixed in Bourn's fluid for
histologic examination and determination of the number of lepto-
tene spermatocytes per Sertoli cell nucleolus. The second testis will
be homogenized and the number of resistant spermatid nuclei
determined. The vesicular glands will be weighed as an indirect
measure of circulating testosterone concentration. The distal half of
one epididymis (half corpus plus cauda) will be weighed. Sperma-
tozoa will be expressed from the severed end of the distal cauda
epididymidis into phosphate-buffered saline containing 0.1% bovine
serum albumin. The percentage of progressively motile spermatozoa
will be determined under phase-contrast microscopy and the inci-
dence of abnormal spermatozoa recorded. A stained slide of the
spermatozoa will be made for documentation. The distal half of the
contralateral epididymis will be isolated, homogenized, and the total
number of sperm heads determined. Concentration of testosterone,
LH, and FSH in serum will be determined.
The females will be killed on day 83 to 89 (18 days after
mating), and the numbers of corpora lutea and implantation sites, as
well as embryo viability, will be determined.
Rabbits. Twelve mature male rabbits (body weight must be
>90% of adult normal for that strain) will be given the >0.5 MTD by
.inhalation, i.p. injection, or in drinking water (the same route of
administration should be used for both rats and rabbits when
-------
84
possible) for a period of 6 cycles of seminiferous epithelium (10.7
days per cycle X 6 = 64 days). An appropriate control group (or
groups) will be established concurrently. Testis size and body weight
will be measured weekly starting on day -14 and continuing through
day 65. Two ejaculates will be collected every 3 to 4 days (e.g.,
Monday and Thursday) using an artificial vagina. The volume, with
and without gel, concentration, and total sperm per ejaculate will be
measured. Sperm motility and gross sperm morphology will be
determined using phase-contrast microscopy. Libido will be sub-
jectively assessed weekly. Each male will be mated with two virgins,
sexually mature females over a 4-day period (between days 54 and
57) and the females allowed to kindle. On day 65, blood will be
taken by cardiac puncture and the male rabbits killed. Evaluation
will be similar to that for rats. Testis size and weight, number of
homogenization resistant spermatids, weight and sperm content of
the distal epididymis, and motility and morphology of sperm from
the distal epididymis will be evaluated. Gross histologic evaluation
and enumeration of the number of leptotene spermatocytes per
Sertoli cell nucleolus will be made on one testis fixed in Bouin's
fluid. Weight of the accessory sex glands will be recorded as an
indirect measure of circulating-testosterone level. Blood will be saved
for possible assay for FSH, LH, and testosterone.
Pregnancy rate and litter size will be determined for females bred
to each male.
Fertility Testing
Tests available
Humans. In vitro oocyte penetration tests are the only means
available for assessing the fertilizing capacity of human sperm. Since
human in vitro fertilization cannot itself be used as a test, substitutes
must be used for the human ovum. These include the zona pellucida
of stored human follicular oocytes (53) and the zona-free hamster
vitellus (54). Regrettably, these tests have not been carefully
validated to establish variation among independent analyses of the
same ejaculate or of different ejaculates from one male. In situations
where in vitro testing of human sperm fertility is indicated, the use
of a double-fluorescent-label competitive sperm penetration assay
with the zona-free hamster egg will increase the sensitivity of the test
(55). The sensitivity of the hamster egg penetration assay is also
increased by attempting to count the total number of sperm per
penetrated hamster vitellus as well as the percentage of penetrated
eggs.
-------
85
Animals. In vivo mating tests should be carried out with
laboratory rats and rabbits. A visible reduction of quality of the
ejaculate may not be reflected in the fertility level because of a
superfluidity of spermatozoa in the ejaculate. The sensitivity of the
test can be increased greatly by insemination with critical numbers of
sperm (ca. 1 X 106 to 2 X 106 in the case of the rabbit). In vitro
tests are unnecessary for use with animals.
Usefulness
The animal tests are useful since they measure the ability of
sperm to reach and fertilize an ovum. The in vivo tests are well
established and reliable if critical sperm numbers are used. Tests of
the fertilizing capacity of human sperm allow assessment of semen
from a human population at reproductive risk when the hazard being
studied has produced fertilization dysfunction in the animal tests. If
fully validated, the in vitro human fertilization system also could be
used to determine the dose-response relationship of a compound to
the fertility of human sperm.
Sensitivity
Although simple mating trials with evaluation of offspring
provide some useful information on fertilization, this is an all-or-
none measurement. Since sperm production is greatly in excess of
that required for fertility, significant reductions in sperm output by
the testes may not be detectable by this method (2). The test can be
improved as an assay of fertilization ability by using artificial
insemination with sperm in limited numbers. Another potentially
useful approach involves competition between two populations of
spermatozoa from males whose status in relation to each other is
known, with expectation of change in the competitive relations
following exposure of the male to an agent of interest (39). This
latter approach requires further validation (see Research Needed).
In fertility tests embryos should be recovered as early as the 2—8
cell stage. Evaluation of a second group of pregnant rats between
days 15 and 19 enables comparison of the numbers of viable and
dead embryos with the number of corpora lutea and is more efficient
than allowing parturition. If rabbit eggs of 2—8 cells are recovered,
the number of sperm associated with fertilized and unfertilized eggs
can be counted. This should bring to light abnormalities in sperm
transport or abnormalities of cleavage that may result from defects in
the sperm genome (see Sperm Nucleus Integrity).
-------
86
Specificity
The fertilization tests are specific indicators of reproductive
toxicity, although they have limited value in the broader context of
toxicological testings. The toxicological end points in the in vivo
animal fertilization system include failures of (a) sperm-egg associa-
tion, (b) sperm penetration of the zona pellucida, and (c) normal
cleavage of the early embryo. All of these events are directly
analogous to those occurring in humans. A consistent failure of
human spermatozoa to penetrate >10% of zona-free hamster eggs
probably reflects an abnormality of the physiological events associ-
ated with fertilization (i.e., sperm capacitation and/or the acrosome
reaction). These events are presumed to be the same as those
required for human fertilization in vivo.
Sperm Nucleus Integrity
A toxic chemical may cause infertility of exposed males through
action on the sperm genome rather than by alteration of the normal
course of spermatogenesis. Thus, the usual parameters used to assess
semen quality will not detect this cause of infertility. Genetic
damage to the spermatozoa is best assessed by mating the exposed
male to untreated females and observing the progeny for sterility,
heritable translocations, sex-chromosome loss, specific locus muta-
tions, mutations affecting the skeleton and eye, and dominant
lethality. These procedures are covered in the U.S. Environmental
Protection Agency's proposed Guidelines for Mutagenicity Risk
Assessment (56). Of these tests, only dominant lethality has a
bearing on the fertility of an exposed male. This effect can be
detected by evaluation of fertilized eggs and embryos as outlined in
the section titled Fertility Testing.
Since the animal tests referred to above are not applicable to the
human male, it would be desirable to be able to assess the genetic
integrity of human spermatozoa directly. Four methods for the
detection of chromosomal abnormalities in spermatozoa are as
follows:
Quinacrine staining for Y-chromosome aneuploidy
This technique, used also by inference for possible somatic
chromosome aneuploidy (57), is easy and economical, does not
require sophisticated equipment, and should be suitable for the study
of population groups. However, the method is subject to many
-------
87
errors, and its accuracy has been questioned (58, 59). The reliability
of the procedure can be improved by using more rigid criteria for
scoring fluorescent Y bodies (60). Nevertheless, the technique's
reliability has not been sufficiently established to warrant its use as a
routine screening procedure.
Spermatozoal morphology
Chemicals, radiation, heat, and a variety of insults increase the
proportion of morphologically abnormal sperm in the ejaculate, and
an increase in the number of abnormal sperm usually results in
impaired fertility. This aspect has been considered in the section on
evaluation of semen quality. That these abnormalities are associated
with chromosomal damage, however, has not been demonstrated.
Karyotyping of human spermatozoa by the
denuded-hamster-egg technique
This procedure is technically difficult, requires highly trained
personnel, and at present should be reserved for evaluation of those
cases where additional evidence that chromosome abnormalities are a
factor in reduced fertility is desired.
Genetic damage with consequences for male fertility also can
arise from strand breaks, base alterations, and base substitution in
the sperm DNA. Except for strand breaks, methods are currently
unavailable for the detection of these lesions. Future work to
develop qualitative and quantitative procedures for the detection of
such lesions in sperm DNA would be desirable, since sperm with an
apparently normal chromosome complement may be responsible for
male fertility problems.
Dose Response
1. The criteria for evaluating male reproductive processes, dis-
cussed above, can be quantified. In most cases, the procedures are
objective, accurate, precise, and sensitive. Data for each criterion
have a considerable response span, although values for normal
individuals may not have a normal distribution.
2. A number of agents acting on the reproductive system are
known to induce a partial suppression in one or more of the criteria
listed when given in low dosages and a more severe effect as the dose
is increased.
-------
88
3. It is likely that separate dose-response curves can be estab-
lished for several criteria with each agent tested. The sensitivity of a
particular test would depend upon the nature of the agent.
4. It is likely that agents affecting reproduction have a threshold
dose below which damage does not occur.
5. It is unlikely, however, that chronic administration of an agent
at >0.5 MTD would not induce a detectable alteration in one or
more of the criteria listed, in at least one of two species, if the agent
in fact has a deleterious effect on reproductive function in the
human male.
6. Reversibility of damage to the male reproductive system often
occurs after exposure to the causative agent is terminated. Complete
regeneration repair usually will require an interval equivalent to at
least three to four and often more than six to twelve cycles of the
seminiferous epithelium.
REFERENCES
1. Amann, R. P.: Sperm, production rates. In: The Testis, Vol. 1, A. Johnson,
W. Gomes, and N. Van Demark, Eds., Academic Press, Inc.: New York; pp.
433-482,1970.
2. Foote, R. H.: Research techniques to study reproductive physiology in the
male. In: Techniques and Procedures in Animal Science Research,
American Society of Animal Producers: Albany, New York; pp. 81-100,
1969.
3. Carson, W. S. and Amann, R. P.: The male rabbit. VI. Effects of ejaculation
and season on testicular size and function. J. Anim. Sci. 34:302—309,
1973.
4. Coulter, G. H. and Foote, R. H.: Bo.vine testicular measurements as
indicators of reproductive performance and their relationship to produc-
tive traits in cattle: a review. Theriogenol. 11:297-311, 1979.
S.Paufler, S. K., Van Vleck, L. D. and Foote, R. H.: Estimation of testicular
size in live rabbits. Int. J. Fertil. 14:188-191,1969.
6. Amann, R. P. and Lambiase, J. T.: The male rabbit. III. Determination of
daily sperm production by means of testicular homogenates. J. Anim. Sci.
28:369-374,1969.
7. Robb, G. W., Amann, R. P. and Killian, G. J.: Daily sperm production and
epididymal sperm reserves of pubertal and adult rats. J. Reprod. Fertil. 54:
103-107,1978.
S.Johnson, L., Petty, C. S. and Neaves, W. B.: The relationship of biopsy
evaluations and testicular measurements to overall daily sperm production
in human testes. Fertil. Steril. 34: 36-40, 1980.
9. Johnson, L., Petty, C. S. and Neaves, W. B.: A new approach to
quantification of spermatogenesis and its application to germinal cell
attrition during human spermiogenesis. Biol. Reprod. 25: 217-226, 1981.
-------
89
10. Mian, T. A., Suzuki, N., Glenn, H. J., Haynie, T. P. and Meistrich, M. L.:
Radiation damage to mouse testis cells from "M technetium per-
technetate. J. Nucl. Med. 18: 1116-1122,1977.
11. Amann, R. P.: A critical review of methods for evaluation of spermato-
genesis from seminal characteristics. J. Androl.; in press, 1981.
12. Lu, C. C., Meistrich, M. L. and Thames, H. O., Jr.: Survival of mouse
testicular stem cells after gamma or neutron irradiation. Radiat. Res. 81:
402-415,1980.
13. Meistrich, M. L., Hunter, N., Suzuki, N., Trostle, P. K. and Withers, H. R.:
Gradual regeneration of mouse testicular stem cells after ionizing
radiation. Radiat. Res. 74: 349-362,1978,
14.Paufler, S. K. and Foote, R. H.: Effect of triethylenemelamine (TEM) and
cadmium chloride on spermatogenesis in rabbits. J. Reprod. Fertil. 19:
309-319,1969.
IS.Swierstra, E. E. and Foote, R. H.: Cytology and kinetics of spermatogenesis
in the rabbit. J. Reprod. Fertil. 5: 309-322, 1963.
16.Clermont, Y.: Kinetics of spermatogenesis in mammals: seminiferous
epithelial cycles and spermatogonial renewal. Physiol. Rev. 52: 198—236,
1972.
17. Berndtson, W. E.: Methods for quantifying mammalian spermatogenesis: a
review. J. Anim. Sci. 44: 818-833,1977.
18. Meistrich, M. L., Reid, B. 0. and Barcellona, W. J.: Changes in sperm nuclei
during spermiogenesis and epididymal maturation. Exp. Cell. Res. 99:
72-78,1976.
19. Jackson, H.: Comparative effects of some antispermatogenic chemicals. In:
The Regulation of Mammalian Reproduction, S.Segel, R. Crozier, P. A.
Corfman, and P.D. Condliffe, Eds., Charles C. Thomas: Springfield,
Illinois; pp. 257-270,1973.
20. Oakberg, C. F.: Gamma-ray sansitivity of spermatogonia of the mouse. J.
Exp. Zool. 134: 343-356, 1957.
21.Parvinen, L. M., Soderstrom, K. 0. and Parvinen, M.: Early effects of
vinblastine and vincristine on the rat spermatogenesis: analyses by a new
transillumination-phase contrast microscopic method. Exp. Pathol. 15(2):
85-96,1978.
22. Homm, R. E., Rustians, C. and Hahn, D. W.: Antispermatogenic effects of
5-Thio-D-glucose in male rats. Biol. Reprod. 17(5): 697-700, 1977.
23.Steinberger, E. and Sud, B. N.: Specific effect of fluoroacetamide on
spermiogenesis. Biol. Reprod. 2: 369-375,1970.
24. Lu, C. C. and Meistrich, M. L.: Cytotoxic effects of chemotherapeutic drugs
on mouse testis cells. Cancer Res. 39: 3575-3582,1979.
25. Withers, H. R., Hunter, N. M., Barkley, H. T. and Reid, B.O.: Radiation
survival and regeneration characteristics of spermatogenic stem cells of
mouse testis. Radiat. Res. 57(1): 88-103, 1974.
26. Courot, M.: Some results obtained in the irradiation with X-rays of testes of
lambs. In: Effects of Ionizing Radiation on the Reproductive System,
W.Carlson and F.Gassner, Eds., Pergammon Press: New York: pp.
279-286,1964.
-------
90
27. Berndtson, W. E.: Methods for quantifying mammalian spermatogenesis: a
review. J. Anim. Sci. 44: 818-833,1977.
28. Steinberger, E. and Tjioe, D. Y.: A method for quantitative analysis of
human seminiferous epithelium. Fertil. Steril. 19: 960-970, 1968.
29. Harasymowycz, J., Ball, L. and Seidel, G. E., Jr.: Evaluation of bovine
spermatozoal morphologic features after staining or fixation. Am. J. Vet.
Res. 37: 1053-1057,1976.
30. Holtz, W. and Foote, R. H.: Composition of rabbit semen and the origin of
several constituents. Biol. Reprod. 18: 286-292,1978.
31-Amann, R. P.: Use of animal models for detecting specific alterations in
reproduction. Appl. Fund. Toxicol.; in press, 1982.
32. Boucher, J. H., Foote, R. H. and Kirk, R. W.: The evaluation of semen
quality in the dog and the effects of frequency of ejaculation upon semen
quality, libido, and depletion of sperm reserves. Cornell Vet. 48(1):
67-86,1958.
33.Amann, R. P.: Effect of ejaculation frequency and breed on semen
characteristics and sperm output of rabbits. J. Reprod. Fertil. 11:
291-293,1966.
34. Freund, M.: Effect of frequency of emission on semen output and an
estimate of daily sperm production in man. J. Reprod. Fertil. 6: 269—286,
1963.
SS.Desjardins, C., Kirton, K. T. and Hafs, H. D.: Sperm output of rabbits at
various ejaculation frequencies and their use in the design of experiments.
J. Reprod. Fertil. 15: 27-32, 1968.
36. Amann, R.P., Kavanaugh, J. F., Griel, L. C., Jr. and Voglmayr, T. K.: Sperm
production of Holstein bulls determined from testicular spermatid
reserves, after cannulation of rete testis or vas deferens, and by daily
ejaculation. J. Dairy Sci. 57: 93-99,1974.
37.Katz, D. F. and Overstreet, J. W.: Sperm motility assessment by video-
micrography. Fertil. Steril. 35(2): 188-193,1981.
38. Overstreet, J. W., Katz, D. F., Hanson, F. W. and Fonseca, J. R.: A simple,
inexpensive method for objective assessment of human sperm movement
characteristics. Fertil. Steril. 31: 162-172, 1979.
39. O'Connor, M. T., Amann, R. P. and Saacke, R. G.: Comparisons of computer
evaluations of spermatozoal motility with standard laboratory tests and
their use for predicting fertility. J. Anim. Sci.; in press, 1981.
40. Seidel, G. E., Jr. and Foote, R. H.: Variance components of semen criteria
from bulls ejaculated frequently and their use in experimental design. J.
Dairy Sci. 56: 399-405,1973.
41. MacLeod, J. and Gold, R. Z.: The male factor in fertility and infertility. II.
Spermatozoan counts in 1000 men of known fertility and in 1000 cases of
infertile marriage. J. Urol. 66: 436-449,1951.
42. Freund, M.: Standards for the rating of human sperm morphology. Int. J.
Fertil. 11:97-180,1966.
43.Belsey, M. A., Eliasson, R., Gallegos, A. J., Moghissi, K.S., Paulsen, C. A.
and Prasad, M. R.N.: Laboratory Manual for the Examination of Human
-------
91
Semen and Semen-Cervical Mucus Interaction, Press Concern: Singapore;
43 pp., 1980.
44. Eliasson, R.: Analyses of semen. In: Progress in Infertility, S. J. Behrman
and R.W. Kistnes, Eds., Little Brown and Co.: Boston; pp. 691-713,
1975.
45. Hammerstedt, R. H.: Monitoring the metabolic rate of germ cells and sperm.
In: Reproductive Processes and Contraception, K.W. McKerns, Ed.,
Plenum Publishing: New York; pp. 353-391, 1981.
46. Thorell, J. I. and Larson, S. M.: Radioimmunoassay and Related Techniques.
C. V. Mosby Co.: St. Louis; 1978.
47. Gomes, W. R.: Chemical agents affecting testicular function and male
fertility. In: The Testis, Vol.3, A.Johnson, W. Gomes, and
N.VanDemark, Eds., Academic Press, Inc.: New York; pp. 483—554,
1970.
48. Jouannet, P., Czyglik, F., David, G., Mayaux, M. J., Spira, A., Moscato, M. L.
and Schwartz, D.: Study of a group of 484 fertile men. I. Distribution of
semen characteristics. Inter. J. Androl. 4: 440—449, 1981.
49. Schwartz, D., Mayaux, M. J., Spira, A., Moscato, M.L., Jouannet, P.,
Czyglik, F. and David, G.: Study of a group of 484 fertile men. II.
Relation between age (20—50) and semen characteristics. Inter. J. Androl.
4: 450-456,1981.
50. Heuchel, E., Schwartz, D. and Price, W.: Within-subject variability and the
importance of abstinence period for sperm count, semen volume, and
pre-freeze and post-thaw motility. Andrologia 13: 479—485, 1981.
5I.Rogers, B. J., Van Campen, H., Veno, M., Lambert, H., Bronson,R. and
Hale, R.: Relationship between morphology and human-sperm fertilizing
ability. Fertil. Stern. 32: 664-670, 1979.
52.Snyder, P. J., Lipschultz, L. I. and Greenberg, S.H.: In: The Testis of
Normal and Infertile Men, P. Troen and H. Nankin, Eds., Raven Press:
New York; 539 pp., 1977.
53. Yanagimachi, R., Lopata, A., Odom, C. B., Bronson, R. A., Malin, C. A. and
Nicolson, G.L.: Retention of biologic characteristics of zona pellucida in
highly concentrated salt solution—use of salt-stored eggs for assessing the
fertilizing-capacity of spermatozoa. Fertil. Steril. 31(5): 562—574, 1979.
54. Yanagimachi, R., Yanagimachi, H. and Rogers, B. J.: Use of zona-free animal
ova as a test-system for assessment of fertilizing-capacity of human
spermatozoa. Biol. Reprod. 15(4): 471-476, 1976.
55..Blazak, W. F., Overstreet, J. W., Katz, D. F. and Hanson, F.W.: A
competitive in vitro assay of human sperm fertilizing ability utilizing
contrasting fluorescent sperm markers. J. Androl; in press, 1982.
56. Federal Register: Proposed guidelines for mutagenicity risks assessments. RP
FRL 1563-2.45(221): 74984-74988, 1980.
57. Kapp, R. W., Jr.: Detection of aneuploidy in human sperm. Environ. Health
Perspect. 31:27-31,1979.
58. Beatty, R. A.: F-bodies as Y-chromosome markers in mature sperm heads: a
quantitative approach. Cytogenet. Cell. Genet. 18: 33-49, 1977.
-------
92
59. Roberts, A. M. and Goodall, H.: Y-chromosome visibility in quinacrine-
stained human spermatozoa. Nature 262: 493—494,1976.
60. Kapp, R. W., Jr. and Jacobson, C. B.: Analysis of human spermatozoa for Y
chromosomal nondisjunction. Teratogen. Carcinogen. Mutagen. 1:
193-211,1980.
-------
II. GLOSSARY OF TERMS USED IN
MALE REPRODUCTION
androgen—a class of steroid hormones produced in the gonads and
adrenal cortex that regulate masculine sexual characteristics; a
generic term for agents that encourage the development of or prevent
changes in male sex characteristics.
backward motility—the movement of a sperm in a reverse direction
(toward the middle piece) rather than a forward direction. Note:
backward motility is typically caused by a 180° reflection of the
middle piece, which may be a secondary abnormality or may be an
artifact induced by temperature shock or osmotic shock.
cellular association or stage—one of a series of characteristic cellular
groupings of different types of germ cells found in a specific area of a
seminiferous tubule. Each association contains several layers of germ
cells, each layer representing one cell generation. These groupings are
not random. Thus, each association contains specific germ cell types
in certain developmental phases. For example, spermatogonia of a
specific type are always found with primary spermatocytes of a
certain developmental phase and spermatids of a certain develop-
mental phase. One cellular association or stage is found at any
moment in a given site within a tubule. Cellular associations are a
consequence of the synchronous evolution of the different germ cell
generations.
circular motility—a clearly discernible motion at a moderate-to-high
velocity, but in circles rather than a more or less linear direction.
cohort of germ cells—all germ cells that are the progency of one
A-spermatogonium. Since cytokinesis is incomplete, all germ cells in
the cohort remain joined by intercellular bridges and develop
synchronously.
cycle of the seminiferous epithelium—the complete series of cellular
associations occurring in the seminiferous epithelium (6 stages in
man; 14 stages in the rat; and generally classified into 8 stages in the
rabbit).
93
-------
94
daily spermatozoal output—the total number of sperm ejaculated
over an interval of at least 7 days after depletion of epididymal
reserves, expressed on a per-day basis. Note: for males ejaculating
once every 1 to 3 days, after the reserves of spermatozoa in the
cauda epididymidis and ductus deferens have been stabilized, daily
spermatozoal output will approach daily spermatozoal production.
daily spermatozoal production—the total number of sperm pro-
duced per day by the two testes.
duration of spermatogenesis—the interval between the time a stem
spermatogonium becomes committed to produce a cohort of
spermatids and the release of the resulting spermatozoa from the
germinal epithelium. It is likely that the duration of spermatogenesis
requires between 4.3 and 4.7 cycles of the seminiferous epithelium
(exact values for most species are unknown). It is difficult to
establish the time interval between formation of the stem spermato-
gonium and formation of preleptotene primary spermatocytes, but
this interval may equal the duration of between 1.2 and 1.7 cycles of
the seminiferous epithelium in many species. Therefore, the term
amputated spermatogenesis is occasionally used to refer to the
portion extending from formation of the preleptotene spermatocytes
through spermiation; this process typically requires about three
cycles of the seminiferous epithelium. The entire duration of
spermatogenesis would total 4.2 to 4.7 cycles of the seminiferous
epithelium (about 72 days in the human and fewer in most animals).
duration of the cycle of the seminiferous epithelium—the interval
required for a cell to pass through one complete series of cellular
associations. This duration is constant for a strain or species (12.9
days for Wistar rat, 10.7 days for rabbit, and 16.0 days for human).
The cycle length is unaffected by environment, hormonal levels, or
cytotoxic damage to the germ cells.
efficiency of spermatozoal production—the number of sperm
produced per day per gram of testicular parenchyma.
ejaculate—the total seminal sample obtained during ejaculation.
ejaculation—the expulsion of semen through the urethra.
-------
95
emission—deposition of sperm and fluids from the caudae epi-
didymidis and ductuli deferentia and fluids from the accessory sex
glands into the pelvic urethra.
flagellating spermatozoon—a sperm (not stuck to the glass slide)
whose position does not change, although its tail moves back and
forth.
follicle-stimulating hormone or FSH—a glycoprotein hormone
secreted by the anterior pituitary of vertebrates that promotes
spermatogenesis and stimulates growth and secretion of the Graafian
follicle.
luteinizing hormone or LH—glycoprotein hormone secreted by the
adenohypophysis of vertebrates that stimulates hormone production
by interstitial cells of gonads.
maximum tolerated dose (MTD)—the highest dose that can be given
during a chronic study without a possibility of shortening an animal's
life other than through its carcinogenicity.
meiosis—two divisions of primary spermatocytes to first form
secondary spermatocytes and secondly to form spermatids. Cells are
called primary or secondary spermatocytes.
nonmotile spermatozoon—a sperm that does not quiver or move a
discernible distance during visual observation.
percentage of motile sperm—the percentage of sperm that are
progressively motile, circularly motile, or backward motile; con-
ventionally estimated as a subjective observation of sperm in a
diluted sample of semen viewed with a phase-contrast microscope.
Note: this percentage can be determined objectively using one of
several procedures.
percentage of progressively motile sperm—the percentage of sperm
that are progressively motile (excluding circularly motile and
backward motile sperm); conventionally estimated as a subjective
observation of sperm in a diluted sample of semen viewed with a
phase-contrast microscope. Note: this percentage can be determined
objectively using one of several procedures.
-------
96
primary abnormality—an abnormality of sperm morphology origi-
nating during spermatogenesis, often associated with the head.
Fertilization of an ovum by the spermatozoon characterized by a
primary abnormality is unlikely.
progressive motility—a clearly discernible, fairly continuous, for-
ward motion at a moderate-to-high velocity in a reasonably linear
path. Progressive motility is greater than 25 /mi/sec for human
sperm. Note: in nonfrozen semen, many sperm will rotate on their
long axis while swimming progressively, although in frozen-thawed
semen, progressive motility may not be accompanied by cellular
rotation. Also, the composition of the buffer used to dilute a sample
of semen can influence whether a motile sperm will rotate or swim
without rotation (flat). Rotation about the long axis and the helical
beat of the tail often move the head of the sperm in a zig-zag path
rather than a true linear path.
quantitative evaluation or objective evaluation—an analytical mea-
surement of sperm motility or velocity performed by a nonbiased
instrument rather than visually by an individual.
quivering spermatozoon—a sperm that rotates slightly on its long
axis or oscillates; characteristic of some sperm recovered from the
efferent ducts or rete testis.
secondary abnormality—an abnormality of sperm morphology
induced during epididymal transit or ejaculation, usually associated
with the tail. When a secondary abnormality is induced in a
spermatozoon, its competitive ability to fertilize an ovum is reduced.
semen—a mixture of sperm and fluids from the excurrent ducts and
accessory sex glands.
seminal volume—the volume of an ejaculate (expressed in milli-
liters).
seminiferous epithelium—the normal cellular components within
the seminiferous tubule consisting of Sertoli cells and germ cells
(spermatogonia, primary spermatocytes, secondary spermatocytes,
and spermatids). Sertoli cells are somatic cells that are usually
nondividing in adult animals and probably are important for
metabolic exchange between the germ cells in the luminal compart-
ment and that, by means of Sertoli-Sertoli junctions, form the
-------
97
blood-testis barrier. They also aid in coordination of spermatogenesis
and have an endocrine function.
spermatogenesis—the sum of the transformations that result in
formation of spermatozoa from spermatogonia and continued
formation of a fairly constant number of uncommitted spermato-
gonia. The entire spermatogenic process is initiated in early
embryonic development and continues after birth and puberty as a
consequence of continual renewal of stem cells. At birth two cell
types are found within the seminiferous tubule: supporting cells,
which give rise to the Sertoli cells of the puberal male, and the
gonocytes, which will develop into spennatogonia. The intense
proliferation of germ cells and the subsequent release of spermatozoa
do not occur randomly. Rather the germinal elements always follow
the same pattern of development (unless particular cells and their
progeny degenerate) within males of a species.
spermatozoal velocity—the velocity with which a progressively
motile or circularly motile sperm moves. Spermatozoal velocity is
conventionally expressed on a subjective scale from 0 (low velocity)
to 4 (maximum velocity) but should be expressed as Mm/sec on the
basis of quantitative measurements.
spermiation—release of spermatozoa from the germinal epithelium
into the lumen of the seminiferous tubule. Prior to release the germ
cells are called spermatids, and after spermiation they are called
spermatozoa.
spermiogenesis—the differentiation of spermatids from spherical
cells with considerable cytoplasm to characteristically shaped cells
with a highly condensed nucleus and scant cytoplasm but with a
flagellum. Cells are called spermatids. Based on changes in the
spermatid acrosome, spermiogenesis can be considered as a
continuum consisting of four phases: Golgi, cap, acrosome, and
maturation. In addition to acrosomal evolution, condensation of the
nuclear material and formation of the flagellum occur.
subjective evaluation—a visual estimate subject to observer bias and
error.
testosterone—a biologically potent androgenic steroid that may be
released from the gonads and adrenal glands.
-------
98
total spermatozoa per ejaculate—the total number of spermatozoa
in an ejaculate (determined as the product of seminal volume times
spermatozoal concentration and expressed as 106). Note: the total
number of sperm per ejaculate, not spermatozoal concentration,
provides the best information on the number of spermatozoa
produced by the testes, since spermatozoal concentration is in-
fluenced by the relative contributions of the accessory sex glands
diluting the bolus(es) of sperm transported during emission from the
ductus deferens and cauda epididymidis.
twitching spermatozoon—a sperm that occasionally or continuously
moves a short distance with a violent motion and then comes to rest,
at least momentarily, before an additional twitch or jump. The
twitch or jump need not be in a forward direction.
-------
CHAPTER 4
CURRENT STATUS OF, AND CONSIDERATIONS
FOR, ESTIMATION OF RISK TO THE HUMAN
CONCEPTUS FROM ENVIRONMENTAL
CHEMICALS
Definition and Scope
Teratology is the study of the causes, mechanisms, and sequelae
of perturbed developmental events in species of animals that undergo
ontogenesis. This report is restricted to a consideration of factors
influencing the current status of risk assessment of teratologic effects
of environmental agents. It is considered a preliminary document
touching upon the major considerations basic to quantitative
estimation of risk to development of the conceptus following
exposure of pregnant animals to environmental agents. This docu-
ment provides no definitive means for assessing risks to the human
conceptus, since no documented or validated system for such
assessment has yet been established. Basic to risk estimation is hazard
assessment, which requires quantification and validation of reliable
end point assays. This document briefly discusses the factors and
scientific considerations upon which degrees of confidence applicable
to contemporary studies of teratology are to be based. Some
additional considerations in evaluating experimental data (e.g., acute
versus chronic exposures) have not been covered explicitly, but
references are provided to aid the reader in gathering further
information.
Impact of Developmental Abnormalities on Humans
Approximately 50% of human conceptuses fail to reach term,
and perhaps as many as half of those lost are structurally abnormal
(1). Approximately 3% of newborn children are found to have one or
more significant congenital malformations at birth, and by the end of
tiie first postnatal year, approximately 3% more (2, 3) are found to
99
-------
100
have developmental malformations. An additional group, whose size
is difficult to estimate, has functional abnormalities of the nervous,
respiratory, gastrointestinal, immunologic, and other systems. Some
unknown proportion of these abnormalities may be due to environ-
mental insult during prenatal life.
Causes of Congenital Malformations
Relatively little is known about the specific causes of most
human congenital defects. It is estimated that 10 to 15% of all
human congenital malformations are due to environmental agents
and another 10 to 15% to hereditary factors (i.e., gene mutations and
chromosomal aberrations). The remainder are considered to result
from unknown causes and from complex interactions between
multifactorially determined hereditary susceptibilities and micro-
environmental factors precipitating abnormal developmental se-
quences within the conceptus and its associated membranes. To date,
only a relatively small number of specific environmental agents and
factors have been identified as causing human malformations (4).
From the above, it is concluded that although regulatory controls
on man-made environmental agents may reduce the incidence of
developmental abnormalities, they will not totally prevent them. It
must be recognized that indications from animal experiments of
adverse effects of environmental agents on development may not
always be corroborated by observations of perturbed development in
human populations. Nevertheless, and in full recognition of these
qualifiers, standard animal testing is presently considered the best
available method for predicting risk of congenital malformation in'
human beings prior to human exposure. Information derived from
such testing can be used to detect and to estimate the magnitude of
hazard posed by specific substances to human prenatal development
and can serve as a basis for estimation of risk.
Qualitative Evaluation of Risk Potential
Interspecies comparisons
Inherent interspecies differences complicate extrapolation of
animal test results to direct determination of human risk. Because a
species identical to the human in all relevant characteristics does not
exist, interspecies differences between human beings and the test
-------
101
species must be considered when data are being evaluated. Interpreta-
tion of these inherent interspecies differences is complicated by
species differences in metabolism and pharmacokinetics of the test
agent and in developmental and other attributes characteristic of the
species. Very little is currently understood about the extent and
nature of the interplay among these many factors as they may affect
the production of a teratogenic event.
Since human beings are manifestly heterogeneous, there is little
doubt that human populations will contain broad degrees of
susceptibility and resistance to the possible adverse prenatal effects
of environmental agents. Because this heterogeneity is largely
determined by genetic variability, it has been reasoned that stocks of
animals bred at random are the most appropriate models for testing
teratogenicity. However, in order to estimate the degrees of
susceptibility that may exist within human populations, both the
average response of the test group and the extent of responses within
it must be considered. This goal can be achieved to some extent by
using several stocks of animals. To make such a procedure even more
sensitive and useful, several inbred strains may also be tested, since
this procedure increases the likelihood that a range of sensitivities
will be uncovered (5). For instance, genetically controlled variations
in embryonic face formation account partly for the sensitivity of
certain mouse strains to spontaneous (6) and teratogen-induced
(7, 8) cleft lip and isolated cleft palate.
Dosing and mode of administration
The test agent should be administered over a range of doses,
including a level sufficient to produce signs of maternal toxicity in
the particular species used. If a teratogenic response is observed, a
dose-response relationship should be determined for the agent and
that specific teratogenic effect. In using test animals, the selection of
dosing intervals must take into account the varying degrees of
sensitivity during organogenesis in that species, the possibility of
enzyme induction or other modifying processes that could result
from repeated administration of the test material, and the practical
aspects of administration that would make the dosing comparable to
that which would likely occur with human exposure.
The route of administration of the test agent may significantly
affect the outcome of an experiment. In general, the route of
exposure for test animals should mimic that of human exposure
where possible, although valuable data may be obtained from other
routes of exposure as well. Differences in the response of a species to
-------
102
the route, dose, and vehicle used for exposure to the test agent may
result in significant variations in blood and tissue levels of the agent
in the maternal and embryo-fetal units (9). These factors may or may
not be of direct significance to teratogenesis, but they must be
recognized as being potentially significant.
Placental transfer
The anatomy and physiology of the placentas of experimental
animals and man present a diverse spectrum of maternal-fetal
connections (10). The chorioallantoic placenta of the human is
approximated by that in some non-human primates, whereas the
common experimental animals have, in addition, a yolk sac placenta,
which also structurally and functionally joins embryo and mother.
The extent to which the yolk sac may supplement or complement
transfer of a previously untested chemical via the chorioallantoic
circulation is largely unpredictable. In most of these species (e.g., rat,
mouse, rabbit, guinea pig), the yolk sac placenta may play a major
role in maternal-fetal exchange of substances during early organo-
genesis. The chorioallantoic placenta, which is readily available for
convenient study at term, is in most cases a totally different
structure from that effecting transfer during the critical stages of
development; therefore great care must be exercised to avoid
unwarranted extrapolation from studies of term chorioallantoic
placenta to presumptions for the function of the two placenta!
structures present earlier in gestation. Lipid solubility, ionic charge,
molecular size, and specific structural configuration all appear to
contribute to the transfer of chemicals between mother and fetus.
Little or no relationship may exist between the embryonic and fetal
concentration of agents and their possible teratologic effect, since
potent teratogens do not always accumulate in the fetus at
concentrations greater than those of agents with low teratogenic
potential. Consequently, increased concentration ratios between the
conceptus and mother do not necessarily allow predictability of
teratogenic or other embryotoxic potential (11, 12). Little is
currently known about the sites of action of teratogenic agents;
therefore, any component of the entire maternal-placental—embryo-
fetal unit and all combinations of such should be considered as
possible site(s) for teratogenic action, until the mechanism of
teratogenic action for given agents is better understood.
-------
103
Pharmacokinetics and metabolism
Other variables that may affect the teratogenicity of an agent in
various species include pharmacokinetics and metabolism and those
exogenous factors that may affect these parameters. Some of the
factors to be considered when evaluating data from these experi-
ments are seasonal and circadian effects on development and
metabolism (13, 14); interaction of pharmacokinetic and placental
hemodynamics (15); possible sites of action of agents; sites of
maternal, placental, and embryonic-fetal metabolism; and deposition
and/or depression or induction of metabolic enzymes (9, 16). Certain
agents may stimulate or inhibit enzyme systems, such as liver
microsomal enzyme systems. In some cases single doses at critical
periods in gestation induce a greater teratologic response than
divided doses on several consecutive days (17-21). Maternal and
embryo-fetal nutritional and endocrine states in various species may
interact with and/or alter metabolism and pharmacokinetics (22), as
may species-specific effects resulting from repeated administration of
the test agent, saturation of metabolizing enzymes, and inhibition or
induction of biotransforming enzymes (16, 23).
Basic to interpreting data from studies of teratology is docu-
mentation of a dose-response relationship and determination of a
treatment level below which adverse effects are not evident in the
data available (no-observed-effect level or NOEL). Threshold levels
may be encountered, and dose levels can exist below which
development of the conceptus suffers no observable deleterious
effect at term. The no-observed-effect level does not guarantee
absolute safety, because uncertainty may result from biological
and/or statistical variation. Failure to detect a deleterious effect on
the end point examined could indicate the absence of a deleterious
effect, but absence of observed effect also could occur if the
magnitude of an effect were below the limit of statistical detection
ability.
Mechanisms of action
The mechanisms underlying abnormal embryonic development
are not well understood. It has proven difficult to determine whether
an observed incidence of abnormal development is the result of an
agent or one of its products acting directly on the conceptus or its
placenta, or if it is achieved indirectly through an initial effect on the
mother. Therefore, the primary site of action by an agent capable of
disrupting development may or may not be the specific malformed
organ and may not even be within the conceptus. Whether or not a
-------
104
chemical is evenly distributed within the mother and conceptus, its
action on a particular tissue or organ may be dependent on cellular
interactions and the particular developmental events characteristic of
specific ontogenetic stages. Increased knowledge of these ontogenetic
events and the interaction of toxic agents or their products with
them is needed both to understand the resulting defects and to
enable better extrapolation of effects seen in one species to
predictions of potential effects in another.
Animal Studies
Standard teratogenicity testing
Schardein (24) has discussed in some detail the current method-
ology and testing approach initially outlined in the 1966 Food',and
Drug Administration's (FDA's) guidelines for reproductive studies
and in the 1967 and 1978 World Health Organization's recommenda-
tions, which are further specified in the U.S. Environmental
Protection Agency's proposed guidelines. Numerous countries have
required studies that are generally similar but that vary in particulars.
The object of the standard protocols is to expose animals to test
materials before breeding of the parental generation, during in utero
development and lactation, and in some instances into adult life of
the offspring. To achieve their goals, the experiments are designed in
three phases or segments, with a multigeneration test for reproduc-
tive effects required in some instances.
The first phase, or Segment-I protocol, calls for dosing of both
male and female animals to begin some calculated time prior to
breeding. Treatment of the young males begins 60 days prior to
breeding, and exposure of the females to the test substance begins
two weeks prior to breeding. Dosing continues for both sexes during
the breeding interval and for the impregnated females throughout
pregnancy and lactation. Other major details of the Segment-I
protocol could be described here, but these will be slighted to
emphasize the basics.
The Segment-I protocol is supposed to examine for possible
adverse effects on estrus; sexual performance; formation of the
gametes; their release from the gonad, transport, and interaction to
form a zygote; and zygote passage to and implantation into the
decidua. Because dosing of the dam continues after mating, the
protocol could reveal effects on placenta! formation and its function;
and because dosing continues throughout gestation and lactation, the
protocol could reveal adverse effects on embryonic or fetal develop-
-------
105
ment and on delivery, nurture, and postnatal development of the
pups. Adverse effects of the test material on the supportive
functional parameters essential for normal occurrence of each of the
above could also become evident. These effects could be as diverse as
effects on food intake or altered endocrine status. .This study is
usually made in rats, and by indicating problem areas, it can serve as
a preliminary to later studies.
The second protocol is oriented more specifically to detect
effects on embryonic development. The study is usually made in
both rats and rabbits. The Segment-II study requires that the males
not be treated with the test compound and that treatment of the
pregnant females not begin until after decidual implantation of the
blastocysts has. occurred. Treatment ceases at the end of major
organogenesis, usually considered as the time of closure of the
secondary palate in the species. Autopsy is performed the day before
expected delivery, when the term fetuses are collected for gross
external examination, after which they are examined for skeletal
development and internal soft-tissue morphology.
The goal of this experiment is to detect adverse effects of a test
material on the developmental events characteristic of major organo-
genesis in the embryo.
A Segment-Ill evaluation is a perinatal and postnatal study
requiring treatment of the dams only. The test agent is administered
during the last third of pregnancy and throughout lactation.
Treatment is not scheduled to begin until after the period of major
embryonic development is completed.
The Segment-Ill safety evaluation was designed to detect adverse
effects of substances on fetal development as well as those
developmental processes that continue into infancy and adolescence.
It is in this study that potential effects on postnatal behavior of the
young are usually evaluated.
A rather elaborate multigeneration protocol is employed for
evaluating selected substances for effects on reproduction over three
generations. The goal of the multigeneration protocol is to reveal
effects caused by accumulated toxicity or by agents effective at low
concentration. These protocols for safety evaluation have had
detailed discussion (14, 25), and for each a data base of considerable
size has accumulated. They are not considered as final, however,
because there is a need for flexibility and exercise of scientific
judgment (26), which could improve their detecting ability in some
instances. There is also some justification for their revision in light of
research findings in related fields since their inception.
-------
106
There are four types of developmental defects: gross anatomical,
death in utero, growth retardation, and functional deficit (4).
Currently, the first tliree of these end points are the only ones that
have a data base sufficient to ensure confidence in their applicability
for use in regulatory decisions. Functional status has been studied
broadly only in recent years and soon may develop end point assays
with specific applicability.
Examination of fetuses to identify gross anatomical defects often
entails judgmental and subjective appraisals based on criteria or
standards established by individual laboratories. The routine test as
performed by many laboratories applying FDA's Good Laboratory
Practices requires highly trained technical and professional personnel.
Even though general standards for defining the limits of normality
and associated terminology have not been established, in general
when selected compounds have been evaluated by various labora-
tories, similar findings have been demonstrated by those using the
routine teratology test and its methods for examining the young
(27, 28).
The overwhelming majority of chemicals known to be terato-
genic in human beings have been demonstrated to be teratogenic in
one or more common laboratory species. Many other agents shown
to be teratogenic in laboratory animals have not yet been docu-
mented as teratogens in humans. The difference may be due to
insufficient epidemiologic data, dissimilarities of exposure levels, or
differences in end points analyzed. A teratogenic response in one
species or strain should be considered indicative of a potential
teratogenic hazard for human beings. However, negative responses in
a few species of experimental animals do not necessarily guarantee
absence of adverse effects in human conceptuses. In the routine
teratology test, no one species has been consistently more predictive.
of human teratogenicity than another species.
Functional teratogenicity testing
Functional alterations may prove to be sensitive indicators of
teratogenic potential. Among those that have been studied following
prenatal exposure, a broad and complex range of behavioral effects
has been described. There is concern that these effects may occur at
doses below those producing gross structural defects or prenatal
death (29). Current literature is based largely on studies of rodents,
particularly rats, in which it has been demonstrated that exposure to
chemicals during periods from early organogenesis through pubes-
cence can result in behavioral impairments. To detect such effects
-------
107
more efficiently, reliable and sensitive test procedures are being
developed in several laboratories (30, 31). Although there are no
agreed-upon testing methods, current studies routinely include the
end points in the following areas: (a) reflex ontogeny, (b) habitation
and reactivity, (c) learning and problem solving, (d) activity level, (e)
motor skills, and (f) sensory processes.
Other functional parameters demonstrated to be affected by
prenatal exposure to chemicals include fertility; reproduction; the
endocrine system; immune competence; xenobiotic metabolism; and
various physiologic parameters, including cardiovascular, renal,
gastrointestinal, respiratory, and hepatic functions (32). Finally, late
sequelae of prenatal exposure to chemicals may be manifested
postnatally as cancers or shortened life span (33, 34).
Gross structural defects or significant growth retardation may
complicate analysis of data from tests of function, and alertness to
potential confounding factors is essential. Permanent changes in
functional systems should be viewed as indicating the potential for
an adverse effect in human beings. Transient changes or delays in
functional ontogeny are still not understood, and their significance
must be further evaluated.
Short-Term Testing Procedures
Prioritizing of chemicals for in-depth study
As a prelude to estimation of potential risks, a series of biological
and informational factors may be applied to a new substance to
possibly trigger further testing to some level in a tier system of
evaluations for teratologic effects. It is considered highly desirable
that substances be prioritized for testing to focus research attention
more readily on substances injuring conceptuses at doses significantly
below those toxic to adults. In attempting to list the factors to be
taken into consideration, the need for short-term systems became
evident because of the large number of evaluations needed and the
fact that many of the available data would be in category 2 below.
Listed below are the factors that, when applicable, would make a
substance a high-priority candidate for further testing. Within each of
the two categories, the factors are listed in decreasing order of
importance.
1. Biological effects data possibly available regarding a substance:
Suspected human teratogenicity; Teratogenicity in domestic
-------
108
animals or wildlife established; Short-term teratology test indi-
cating a significant developmental hazard potential; Adult toxic
dose/developmental toxic dose ratio large; Toxicity documented
in the adult at low dosage;
2. Additional information available regarding a substance: Large
numbers of women exposed; Bioaccumulation evident; Persis-
tence of substance in environment; New substance; Involuntary
exposure.
The number of agents in use and potentially impinging on human
development is already vast and is increasing rapidly. Some unknown
small fraction of these may be potentially harmful. to human
conceptuses at doses below those obviously deleterious to adults.
Short-duration and low-cost methods for detecting and prioritizing
those substances posing the greatest potential hazard to the
conceptus are needed. Because standard tests in animals are quite
costly, only a rather small number of substances of potential
teratogenic risk can be evaluated each year. This situation requires
development and validation of short-term methods that will permit
rapid and meaningful testing of these chemicals. It is necessary to
develop, validate, and use assays that will permit more economical
testing of a larger number of agents than could be tested rapidly and
conveniently by standard teratogenicity evaluations. Validation
should consist of various forms of positive correlation between the
results of such tests and those found in conventional in vivo test
procedures. Particular attention should be directed to correlations
with known human' teratogenic responses whenever reliable data are
available. Included should be chemicals already known to have
significant hazard potential for the conceptus, as well as chemicals
considered as lacking such potential. If properly validated short-term
tests were to indicate either potential hazard or safety, such
determinations would be helpful for establishing a priority system
for further tests aimed toward quantitative risk assessment.
Characteristics of short-term assays
Short-term tests might be considered as either preliminaries to
more detailed evaluations, or they might serve as efficient means to
detect those substances capable of posing the greatest hazard to the
conceptus. Whether they serve either or both of these slightly
different goals, the tests should possess certain attributes: they
should be rapid, economical, and reproducible from one laboratory
to another; should have easily identifiable end points; and ideally
-------
109
should prioritize substances according to their potential for posing
hazards to the conceptus. Confidence in their applicability would be
increased by demonstration of dose-response relationships. They
should give minimal false negatives; it is understood that false
positives can be explored further in more elaborate animal testing.
Ideally, the system should encompass as many as possible of the
developmental events known to occur in the conceptus.
Short-term tests may serve as preliminary screens to aid in the
detection of possibly teratogenic hazards. To accomplish this, several
such tests would probably have to be performed concurrently or
sequentially. It must be remembered that such indications of
teratogenic hazard potential must be used prudently for estimating
risk to human beings. Risk estimation can only be achieved by the
use of systems that have been extensively validated, and to date, only
the more routine standard testing methods are considered applicable
to this use.
Potential short-term systems
In vivo. Two in vivo systems have been advanced as possible
short-term systems. One is an abbreviated version of a standard
teratology test protocol using the maternal maximum tolerated dose
and neonatal evaluation shortly after birth (35). The second is
evaluation of homeotic shifts that may prove effective for detecting
minimal expression of teratogenic hazard potential (36). Each system
has potential merit, but as in the in vitro systems, needs careful peer
review or detailed validation.
In vitro. Artificial invertebrate "embryos," embryonic insect
cells, amphibian and fish embryos, or organs of avian and mammalian
embryos (palatal shelves, tooth bud, kidney mesenchyme, pancreas,
bone primordia, lens, sex organs, etc.) have the potential to serve as
the basis for in vitro systems. Table 12 lists a number of potential in
vitro systems. In these procedures, cells, organs, or whole embryos
have been exposed to various chemicals and their effects measured
with the end point permitted by each system. When attempted in a
few instances, adequate dose-response relationships were obtained by
some systems. In their present state, the systems listed in Table 12
have been only partially validated (36-50), and their closer
examination is necessary.
-------
110
TABLE 12 Some/n Vitro Short-Term Systems
Currently in Various Stages of Development
System
Invertebrates
Fish
Vertebrate cell
culture
Organ culture
Whole embryo culture
Developmental
parameters
monitored References
Hydra
Planaiia
Drosophila cells
Zebra fish
Chick embryo
neural crest
Chick embryo
limb bud
mesenchyme
Mouse tumor cells
Terato carcinoma
stem cells
Mouse embryo
limb bud
Rat, mouse,
chick
Various
Regeneration,
dose-response
relationship
of developmental
toxicity
Differentiation
Dose-response
relationship of
developmental
toxicity
Morphology,
differentiation
Differentiation
Cell attachment
Differentiation
(in vivo or
in vitro)
Growth, dysmo'rpho-
genesis,
differentiation
Growth, dysmorpho-
genesis,
histogenesis
36
37,38
39
40
41
41
42
43
44,45
46,47
16,48
49,50
Quantitative Risk Assessment
Quantitative risk assessment is based on the relationship between
laboratory findings and expected human response. If an agent
demonstrates teratogenicity in any mammalian species, some concern
about prenatal human exposure to the agent is justified. The level of
concern is to be tempered by numerous considerations, not the least
of which is the extent of maternal toxicity evident at the dose level
needed to elicit a toxic response in the conceptus (51). It is assumed
that margins of safety applied to the experimental data in test species
can be used to estimate an allowable exposure in pregnant women. It
is considered that no-observable-effect and/or threshold levels of
-------
Ill
exposure do exist (52) for at least some teratogens. The determina-
tion of human risk requires the definition of moderating (or
modulating) conditions such as the distribution of the compound
within the environment; its pattern of use and exposure (whether
intermittent or chronic); and the identification of those subpopula-
tions that may be at high risk as a result of factors such as life style,
age, occupation, etc.
The use of animal test systems under highly controlled experi-
mental conditions has conditional validity for defining human risk.
Although only warning systems at best, laboratory experiments can
provide, in addition to the factors already mentioned, two types of
information that may be useful for estimation of the potential
human risk. These are (a) the ratio of the adult and developmental
toxic doses and (b) the shape of the curve of the teratogenic dose
response. Although not markedly informative to date, more detailed
delineation of projected effects may be obtained through use of
pharmacokinetic information, focusing on access of the agent to
relevant site(s) of teratogenic action (53). However, knowledge of
teratogenic mechanisms and identification of the sites actually
relevant must be obtained before these considerations can achieve
their full potential utility.
It is often necessary to conduct animal experiments at dosage
levels exceeding estimated levels of human exposure to increase the
likelihood that a weak teratogen will produce an apparent effect and
to compensate for the relatively small numbers of animals used in the
test. This requires extrapolation of results from experimental dosage
levels to lower levels of human exposure. There is no uniform basis
for selecting the appropriate mathematical model for such extrapola-
tion.
Safety factors may be applied to establish acceptable dosage
levels that are expected to yield acceptable levels of risk. The size of
the safety factor depends upon the quality and quantity of the
biological effects data available.
For many biological systems, the dose-response curve tends to
flatten at low doses, and for some teratogens this is an important
consideration (54). Hence, decreasing dosage by a safety factor of F
will generally decrease risk by more than a factor of F. That is, if the
upper confidence limit on the risk is U at an experimental dosage of
d, the potential risk at a lower dosage of d/F is predicted to be less
than U/F for the test animal. The uniformity with which this would
apply to potential teratogenic hazards is undetermined as is the
degree of interspecies uniformity for the difference between the
adult (A) toxic and developmental (D) toxic doses. Presence or
-------
112
absence of uniformity in the A/D ratio or slope of the dose-response
between studies in different species would also influence the
magnitude of safety factors.
Priorities for Future Research in Teratology
The areas of research in teratology recommended below focus on
two broad objectives: (a) development of practical and informative
testing systems with which to evaluate both the plethora of
chemicals now in existence and those yet to be developed and (b)
scientific advancement in teratology so that the currently employed
and largely standard test systems can become more useful and
reliable for human risk estimation. The second objective does not
imply that currently available methods cannot be used for human
risk estimation. An opposite view is held, and within limits, such
estimations are possible at the present time on the basis of data from
current state-of-the-art studies. It is felt, however, that methods are
needed to identify more rapidly those chemicals potentially the most
hazardous and to expand the understanding and applicability of all
test methods.
1. The degree to which the end point determinations of adverse
effects on development encountered in the standard protocols
predict adverse effects in other species (especially humans) has not
been reported in detail sufficient for precise quantification of human
risk. Such studies are encouraged as are those that may indicate how
the predictive ability of tests in animals can more precisely herald
human responses.
2. Validated methods are needed for rapid and inexpensive
detection of substances uniquely toxic to conceptuses (i.e., sub-
stances that are embryotoxic at doses below those producing adult
toxicity).
3. A better understanding of mechanisms of teratogenic action or
elucidation of the steps in the pathway between exposure and effect
might significantly improve and refine end point assays. Similar
studies of pathogenesis are needed for effects on biochemical and
physiological systems in the dam, uterus, and placenta.
4. Development of a broader data base on comparative metabo-
lism and pharmacokinetics correlated with teratologic end points
may eventually enhance ability to make interspecies extrapolations.
5. There is a need to develop and validate methods to detect and
quantify possible functional impairments.
6. Methods are needed to detect and predict additive or
synergistic effects more effectively.
-------
REFERENCES
l.Hertig, A. T.: The overall problem in man. In: Comparative Aspects of
Reproductive Failure, K. Benirschke, Ed., Springer-Verlag: New York; pp.
11-41,1967.
2. McKeown, T. and Record, R. C.: Malformations in a population observed for
five years after birth. In: Ciba Foundation Symposium on Congenital
Malformations, G.E.W. Welstenholme and C.M. O'Conner, Eds., Little
Brown: Boston; pp. 2—16,1963.
S.Mellon, G. W. and Kalzenstein, M.: Increased incidence of malformations —
chance or change? J. Am. Med. Assoc. 187: 570-573,1964.
4. Wilson, J. G.: Environment and Birth Defects. Academic Press: New York;
305 pp., 1973.
S.Kalter, H.: Interplay of intrinsic and extrinsic factors. In: Teratology:
Principles and Procedures, J. G. Wilson and J. Warkany, Eds., University of
Chicago Press: Chicago; pp. 57-80, 1965.
6.Trasler, D. G.: Pathogenesis of cleft lip and its relation to embryonic face
shape in A/J and C57BL mice. Teratology 1: 33-49, 1968.
7.Trasler, D. G.: Aspirin-induced cleft lip and other malformations in mice.
Lancet 1: 606-607, 1965.
S.Trasler, D. G. and Fraser, F. C.: Time-position relationships with particular
reference to cleft lip and cleft palate. In: Handbook of Teratology, Vol. 2,
J.G. Wilson and F.C. Fraser, Eds., Plenum Press: New York; pp.
271-292, 1977.
9. Larsson, K. S.: Contributions of teratology to fetal pharmacology. In: Fetal
Pharmacology, L. O. Boreus, Ed., Raven Press: New York; pp. 401—418,
1973.
10. Beck, F.: Comparative placental morphology and function. Environ. Health
Perspect. 18: 5-12,1976.
ll.Waddell, W. J. and Marlowe, G. C.: Disposition of drugs in the fetus. In:
Perinatal Pharmacology and Therapeutics, B. L. Mirkin, Ed., Academic
Press: New York; pp. 119-268,1976.
12. Wilson, J. G., Scott, W. J., Ritter, E. J. and Fradkin, R.: Comparative
distribution and embryotoxicity of methotrexate in pregnant rats and
rhesus monkeys. Teratology 19: 71-80,1979.
13.Barr, M., Jr.: Prenatal growth of Wistar rats: circadian periodicity of fetal
growth late in gestation. Teratology 7: 283-288, 1973.
14. Layton, W. M.: An analysis of teratogenic testing procedures. In: Congenital
Defects, New Directions in Research, D. T. Janerich, R. G. Skalko and
I. H. Porter, Eds., Academic Press: New York; pp. 205-217,1974.
15.Wolkowski-Tyl, R. M.: Strain and tissue differences in cadmium-binding
protein in cadmium-treated mice. In: Developmental Toxicology of
Energy-related Pollutants, D. D. Mahlum, M. R. Sikov, P. L. Hackett and
F. D. Andrew, Eds., Technical Information Center: Oak Ridge, Tennessee;
pp.568-585,1978.
-------
114
16.Fantel, A. G., Greenaway, J. C., Juchau, M. R. and Shepard, T. H.:
Teratogenic bioactivation of cyclophosphamide in vitro. Life Sci. 25:
67-72,1979.
17. Russell, L. B., Badgett, S. K. and Saylors, C. L.: Comparison of the effects of
acute, continuous, and fractionated irradiation during embryonic devel-
opment. Int. J.Radia. Biol.,Suppl: 343-359, 1960.
18. Wilson, J. G.: Effects of acute and chronic treatment with Actinomycin D
on pregnancy and the fetus in the rat. Harper Hosp. Bull. 24: 109—118,
1966.
19. King, C. T. G., Horigan, E. and Wilk, A. K.: Fetal outcome from prolonged
versus acute drug administration in the pregnant rat. In: Drugs and Fetal
Development, M. A. Klingberg, A. A. Abramovici and J. Chemke, Eds.,
Plenum Press: New York; pp. 61-75,1972.
20. Tuchmann-Duplessis, H.: Teratogenic Action of Drugs. Pergammon Press:
New York; 1965.
21.Belisle, R. J. and Long, S. Y.: Tolbutamide treatment of pregnant mice:
repeated administration reduces fetal lethality. Teratology 13: 65—70,
1976.
22.Neubert, D., Merker, H.J., Kohler, E., Krowke, R. and Barrach, H.J.:
Biochemical aspects of teratology. In: Advances in the Biosciences,
G.Raspe, Ed., Pergammon Press: Oxford; pp. 575-622, 1971.
23. Bakay, B. and Nyhan, W; L.: Effects of Thalidomide and Chlorcyclizine on
the biosynthesis of nucleic acids and proteins in fetal and maternal tissue
of therat.J.Pharmacol.Exp.Ther. 171(1): 109-117,1970.
24. Schardein, J. L.: Drugs as Teratogens. CRC Press: Cleveland; pp. 9—12,
1976.
25.Kelsey, F. O.: Present guidelines for teratogenicity studies in experimental
animals. In: Congenital Defects, New Directions in Research, D.T.
Janerich, R. G. Skalko and I. H. Porter, Eds., Academic Press: New York;
pp. 195-204,1974.
26. Golberg, L. M. B.: Discussion pp. 53-55. In: Methods for Detection of
Environmental Agents that Produce Congenital Defects, T.H. Shepard,
J. R. Miller, and M. Marois, Eds., North-Holland Publishing Co.: Amster-
dam; 1975.
27. Wilson, J. G.: Methods for administering agents and detecting malformations
in experimental animals. In: Teratology: Principles and Techniques, J. G.
Wilson and J.Warkany, Eds., University of Chicago Press: Chicago; pp.
262-277,1965.
28. Staples, R. E. and Schnell, V. L.: Refinement in rapid clearing technique in
the KOH-alizarin-red-S method for fetal bone. Stain Technol. 39: 61—63,
1963.
29. Hutchings, D. E.: Behavioral teratology: embryopathic and behavioral
effects of drugs during pregnancy. In: Studies on the Development of
Behavior and the Nervous System, Vol. 4, Early Influences, G. Gottlieb,
Ed., Academic Press: New York; pp. 7—34, 1978.
-------
115
30. Vorhees, C. V., Brunner, R. L. and Butcher, R. E.: Psychotropic drugs as
behavioral teratogens. Science 205: 1220-1225, 1979.
31.Buelke-Sam, J. and Kimmel, C. A.: Development and standardization of
screening methods for behavioral teratology. Teratology 20: 17-29, 1979.
32. Kimmel, C. A.: A profile of developmental toxicity. In: Developmental
Toxicology, C. A. Kimmel and J. Buelke-Sam, Eds., Raven Press; in press,
1981.
33. Rice, J. M.: Perinatal period and pregnancy: intervals of high risk for
chemical carcinogens. Environ. Health Perspect. 29: 23-27,1979.
34. Spyker, J. M.: Assessing the impact of low-level chemicals on development:
behavioral and latent effects. Fed. Proc. Fed. Am. Soc. Exp. Biol. 34:
1835-1844,1975.
35.Chernoff, N. and Kavlock, R. J.: A potential in vivo screen for the
determination of teratogenic effects in mammals. Teratology 21:
33A-34A, 1980.
36. Johnson, E. M.: A subvertebrate system for rapid determination of potential
teratogenic hazards. J. Environ. Pathol. Toxicol. 4: 153—156, 1980.
37. Best, J. B., Morita, M., R.agin, J. and Best, J., Jr.: Acute toxic responses of
the freshwater planarian, Dugesia dqrotocephala, to methyl-mercury. Bull.
Environ. Contarn. Toxicol.; in press, 1981.
38. Best, J. B., Morita, M. and Abbotts, B.: Acute toxic responses of the
freshwater planarian, Dugesia dorotocephala, to chlordane. Bull. Environ.
Contam. Toxicol.; in press, 1981.
39. Bournias-Vardiabasis, N., Terplitz, R. L. and Seecof, R. L.: An in vitro assay
for teratogenesis. Teratology 21: 29A, 1980.
40. Streisinger, G.: Invited discussion on the possible use of zebra fish for the
screening of teratogens. In: Methods for Detection of Environmental
Agents that Produce Congenital Defects, T. H. Shepard, J.R. Miller and
M.Marois, Eds., North-Holland Publishing Co.: Amsterdam; pp. 59—61,
1975.
41. Wilk, A. L., Greenberg, J. H., Horigan, E. A., Pratt, R. M. and Martin, G. R.:
Detection of teratogenic compounds using differentiating embryonic cells
in culture. In Vitro 16: 269-276, 1980.
42. Braun, A. G., Emerson, D. J. and Nichinson, B. B.: Teratogenic drugs inhibit
tumor cell attachment to lectin-coated surfaces. Nature 282: 507—509,
1979.
43. Filler, R.: An in vitro/in vivo coupled prescreen to identify teratogens
requiring metabolic activation. Teratology 21: 37A, 1980.
44. Kochhar, D. M.: The use of in vitro procedures in teratology. Teratology 11:
273-288, 1975.
45.Neubert, D. and Barrach, H. J.: Significance of in vitro techniques for the
evaluation of embryotoxic effects. In: Methods in Prenatal Toxicology:
Evaluation of Embryotoxic Effects in Experimental Animals, D. Neubert,
H. J. Merker and T. E. Kwasigroch, Eds., Georg Thieme: Stuttgart; pp.
202-209,1977.
-------
16
46. Manson, J. M. and Simons, C. R.: In vitro metabolism of cyclophosphamide
in limb bud culture. Teratology 19: 149-158, 1977.
47.Kochhar, D. M. and Agnish, N. D.: Teratogenic testing in vitro. In: Toxicity
Testing In Vitro, R. M. Nardone, Ed., Academic Press: New York; in press,
1981.
48. New, D. A. T.: Techniques for assessment of teratologic effects: embryo
culture. Environ. Health Perspect. 18: 105-110, 1976.
49. Brown, N. A., Goulding, E. H. and Fabro, S.: Ethanol embryotoxicity:
direct effects on mammalian embryos in vitro. Science 206: 573—575,
1979.
50. Klein, N. W., Vogler, M. A., Chatot, C. L. and Pierro, L.J.: The use of
cultured rat embryos to evaluate the teratogenic activity of serum:
cadmium and cyclophosphamide. Teratology 21: 199-208,1980.
51. Johnson, E. M.: Screening for teratogenic potential: are we asking the proper
question? Teratology 21: 259,1980.
52. Staples, R. L.: Teratogens and the Delaney Clause. Science 185: 813, 1974.
53. Young, J. F. and Holson, J. F.: Utility of pharmacokinetics in designing
toxicological protocols and improving interspecies extrapolation. J.
Environ. Pathol. Toxicol. 2: 169-186,1978.
54.Jusko, W. J.: Pharmacodynamic principles in chemical teratology: dose-
effect relationship. J. Pharmacol. Exp. Ther. 183: 469-480, 1972.
-------
CHAPTER 5
••«M^HI^
OTHER CONSIDERATIONS:
EPIDEMIOLOGY, PHARMACOKINETICS,
AND SEXUAL BEHAVIOR
Epidemiology: Methods and Limitations
Epidemiology has been defined as the study of the distribution
and determinants of disease and injury in humans. It focuses on the
occurrence of disease in groups of individuals or populations rather
than in any single individual (1). Ideally, reproductive and teratologic
effects of environmental agents would be assessed in epidemiologic
studies of human populations because of the difficulties inherent in
extrapolating from other species. However, ethical considerations
render randomized controlled trials generally unfeasible. If one
cannot experiment, then one can only observe, but in some
circumstances even observation is not possible (e.g., risk assessment
of new chemicals before they are introduced into the human
environment).
Observational epidemiologic studies can be classified into those
that generate hypotheses and those that formally test hypotheses and
quantify risks. The following list is not an exhaustive delineation of
all possible epidemiologic approaches in these categories, but it
includes those which may be most useful for environmental risk
assessment.
Hypothesis—generating studies
Case reports are a source for raising suspicions about substances
that might be teratogens or reproductive hazards (e.g., an astute
clinician's association of thalidomide with phocomelia). However,
only very striking or very rare outcomes can be detected in this
manner and often after a considerable length of time. For the vast
majority of pregnancy outcomes, formal epidemiologic studies
involving appropriate comparison groups are required.
-------
118
Correlational studies evaluate the patterns of morbidity or
mortality in populations where classification is made on the basis of
aggregates of individuals as distinct from single individuals (e.g.,
geographic-specific spontaneous abortion rates correlated with area-
specific air pollution levels).
In demographic studies, routinely collected information is used
to estimate disease rates in populations composed of individuals
classified by limited demographic characteristics (e.g., age and sex),
allowing for the identification of subgroups at particularly high risk
and of changes in the rates over time.
Population-based registries can detect changes in the incidence of
the outcomes being registered (e.g., spontaneous abortions, low birth
weight, birth defects, neonatal deaths). If placed in selected areas of
suspected high risk and "clean" areas of presumed low risk, they may
point up differences potentially resulting from environmental causes.
Their case materials are a useful resource for mounting case-control
studies of suspected environmental hazards (see below).
Analytic studies for formally testing hypotheses and
quantifying risks
In the case-control design, a series of individuals with an observed
effect (the outcome of interest) and a series of unaffected individuals
are compared with respect to their previous exposure to the
environmental agent of interest. The case-control method is most
appropriate for the study of extremely rare outcomes, such as
ambiguous genitalia and specific birth defects, and relatively rare
outcomes, such as infertility and ectopic pregnancy. It is not useful
for the study of very rare exposure unless the study is conducted in a
selected setting with sufficiently large numbers of exposed persons
(e.g., occupational settings). Case-control studies are not feasible
when previous exposures cannot be ascertained.
In cohort studies, cohorts of exposed and unexposed individuals
are followed for the subsequent occurrence of the outcomes of
interest. This design may be the most desirable method for studying
fairly common outcomes (e.g., spontaneous abortions), for deter-
mining conception rates, and for evaluating subtle indications of
reduced fertility, such as variations in interpregnancy interval.
Limitations
All observational studies are subject to certain limitations.
Studies classified as hypothesis-generating lack information on
-------
119
potential distorting factors; for this reason, among others, they
cannot be used to establish cause-effect relationships. Formal
analytic studies do have this capacity if they are valid (i.e., relatively
free of biases attributable to selection, measurement, and con-
founding).
If losses to follow-up are related to outcome status (cohort
studies) or if entry into the study is related to exposure status
(case-control studies), then selection bias will be present. This can
occur in industrial settings, for example, if individuals who are
exposed to hazardous substances tend to leave the industry and
become lost to follow-up because they become ill.
Errors in measurement of exposure or outcome, if unequal
between the groups being compared, can lead to overestimation or
underestimation of an effect. Equal measurement errors will always
lead to attenuation of an effect, and this is a particular problem
when exposure or outcome is difficult to measure, as is the case with
many environmental exposures and some reproductive outcomes. In
cohort studies, there is particular concern that the ascertainment of
subsequent outcome be unbiased, while in case-control studies, there
is particular concern that the measurement of prior exposure to the
agent of interest be unbiased.
Because observational epidemiologic studies deal with non-
randomized populations, a central concern is whether the groups
being compared are similar in all relevant characteristics. If they
differ in factors related to both the exposure and the outcome, then
confounding bias will be present. Properly conducted epidemiologic
studies will make allowance for all known risk factors of the health
outcome of interest, either in the design or in the analysis.
For assessment of the validity of a particular study, detailed
information about recruitment and participation of the study
population, measurement of the parameters of interest, method of
analysis, and efforts to assess potential biases must be available. A
single epidemiologic study, even if valid, can seldom by itself rule out
chance or bias as the explanation of an observed association.
Establishment of causal associations usually requires the accumula-
tion of consistent evidence from valid studies of human populations.
That cause precedes effect must also be demonstrated. Belief in a
particular hypothesis can be strengthened by evidence from animal
studies, by a biologically plausible mechanism, and by a dose-
response relationship. In general, the stronger an association, the
easier it is to establish a causal association. Conversely, the smaller
the effect, the more difficult it is to demonstrate. Failure to detect
-------
120
an effect may simply reflect inadequate sample size or insufficient
time for the outcome to become manifest.
This brief review touches on only a few of the potential
limitations of observational epidemiologic studies. The assessment of
the validity of any particular study requires extensive knowledge of
epidemiologic methods and experience with their application.
Possible data sources and useful approaches
Currently few epidemiologic studies attempt to detect human
teratogenic and reproductive hazards or to quantify their effects.
Furthermore, there is no systematic application of epidemiologic
methods for this purpose. It would be desirable to have programs
specifically designed to raise suspicions and to test hypotheses. To be
effective, these programs must be supported on an ongoing basis.
Before pilot testing any new epidemiologic program, however,
the potential usefulness of existing studies and data bases should first
be evaluated. Several examples of potentially useful systems are given
below.
There are several registries of birth defects in the United States,
for example, the Birth Defects Monitoring Program of the Center of
Disease Control, which collects information from selected hospitals
throughout the country, and the birth defects registries of metro-
politan Atlanta and of Nebraska and Florida. The development in
selected regions of population-based registries of reproductive health
outcomes (including birth defects, ectopic pregnancies, and sponta-
neous abortions) could point to potential environmental hazards.
Even in a particular geographic area, clusters of cases or changes in
rates over time could suggest a source of environmental contamina-
tion.
Vital statistics have been analyzed from time to time, depending
upon the interest of the investigator. For example, infant mortality
rates have correlated with chlorination levels of public water supplies
in New York (1). In addition, vital statistics have been used as
indicators of reduced male fertility in an occupational setting (2, 3).
A systematic ongoing analysis of vital statistics data in relation to
routinely collected environmental data might be quite useful for
raising suspicions about environmental hazards.
Two ongoing surveillance systems based on the case-control
approach are currently in operation: one is designed to discover and
to evaluate adverse drug effects that are serious enough to warrant
hospitalization (4, 5), arid the second is directed to the discovery and
-------
121
evaluation of drug teratogenic effects (6). In principle this meth-
odology is applicable to the discovery of environmental agents that
are reproductive hazards or teratogens. For the surveillance of
occupational exposures, programs could be located in specified areas
of the country where occupational exposures to suspected hazards
are high. The application of this methodology to the study of
nonoccupational environmental exposures is more difficult, in part
because individuals may not be aware of what they have been
exposed to.
The cohort method has been used to identify several health
hazards in occupational cohorts (7, 8). Other cohorts that might be
useful are enrollees hi health maintenance organizations (HMOs) (see,
for example, Ref. 9). However, an HMO data base has the limitation
that only outcomes that come to medical attention can be studied.
Moreover, no HMO currently has a computerized data base in a form
that would be useful for the conduct of epidemiologic studies of
environmental exposures. A large investment would be required to
build and maintain such a data base.
Pharmacokinetics
Pharmacokinetics can be defined as the quantitative study of the
absorption, disposition, metabolism, and elimination of drugs,
poisons, and other chemical agents in the body. It is important to
evaluate pharmacokinetic variables at different doses and routes of
exposure to understand the toxicological significance of exposure.
Pharmacokinetics can be employed for at least two purposes:
definition of the concentration levels of the agent in blood or in
tissues where the site of action is presumably located, and quantita-
tive description and prediction of the relevant concentration levels,
usually with a mathematical model. The first should be routinely
done to the extent possible as an aid in interpreting other
measurements being made. The second requires much more compre-
hensive study but has the potential for predictive purposes such as
risk assessment.
At this time, there are several basic textbooks of pharma-
cokinetics: Notari (10) presents an elementary overview, but with,
many applications; Wagner (11) and Gibaldi and Perrier (12) provide
collections of the basic mathematical models and solutions with
illustrations of their use. These classical treatments permit organiza-
tion of pharmacokinetic data, along with some biological interpreta-
tions of amounts of an agent in the "central" regions of the body
-------
122
(blood, vital organs) versus "peripheral" regions (other tissues).
However, for use of measured levels in specific tissues, models
incorporating what is known about quantitative aspects of anatomy
and physiology have been found useful (13, 14). Another important
feature of this alternative approach is to enable use of known
physiological and pharmacological differences between animal
species to define some of the critical parameters for quantitative
extrapolation to man. A review is given by Dedrick (15), and
suggestions for defining similarities between animal species are
described by Dedrick and Bischoff (16).
A survey of application of the above pharmacokinetic approaches
to some areas of toxicology is given by Gehring et al. (17), and
further discussions are in chapters of World Health Organization
Environmental Health Criteria (18) and Filov et al. (19). Some
specific issues of importance concerning reproductive and teratogenic
effects are described by Young and Holson (20).
When pharmacokinetics is applied to the specific area of
teratology, the major determinants of the teratogenic agent's
reaching and accumulating in the conceptus are the usual aspects of
pharmacokinetics in the mother, plus the unique features of
transplacental transport, and pharmacokinetics in the conceptus. The
maternal pharmacokinetics may be monitored by the blood half-life
(20), although it may also be desirable to have more complete details
of the disposition into the uterine tissue, as well as the presence of
active metabolites and inducible catabolic enzymes, later mobiliza-
tion of stored agent, and any differences between pharmacokinetics
in chronic versus acute exposures. The uptake and disposition into
the conceptus may be partially predicted from knowledge of
placental membrane transport parameters (models of oxygen and
glucose transport may be a useful basis [22]), specific and
nonspecific binding, and other features of the developing fetus. It is
crucial to determine these effects during the period of major
organogenesis.
Few available studies have applied pharmacokinetic principles to
specific areas of female or male reproductive organs, especially with
reference to formulating models that could be used for predictive
purposes. In one of the few, uptake of cancer chemotherapeutic
agents into the human uterus has been successfully described by a
physiological pharmacokinetic model, which was then used to
formulate clinical dosage regimens (22). In another, Lee and Dixon
(23) present the results of their innovative study of the pharma-
cokinetic determinants of uptake into male gonads.
-------
123
Clearly, much more needs to be done before pharmacokinetics
can be routinely utilized as an adjunct in better defining the basis for
risk assessment of reproductive and teratogenic toxic effects.
However, information obtained using the reasonably well developed
methods described in the earlier references should aid in developing
methods to resolve some of the issues.
Sexual Behavior
Introduction
Overview. The behavioral aspects of reproduction encompass a
broad spectrum of activities including courtship behavior, sexual
behavior, parental behavior, and a variety of social activities that
subtly influence the probability of reproductive success. The scope
of this discussion is limited primarily to sexual behavior for the
following reasons.
• This behavioral aspect of reproduction has received the most
detailed and extensive attention from clinicians and laboratory
investigators.
• Evaluation procedures for sexual behavior of a variety of
animal species are well established.
• Choices can be made among several existing standardized
procedures currently in use in laboratories around the country.
• Observational methods are simple and direct, and workers at a
moderate level of skill can be quickly trained to obtain reliable
measurements.
• Many of the testing procedures recommended in the following
presentation can provide insight into the probable locus of
action of the putative toxicant, and this could not be as easily
achieved if the scope of the investigations was extended to
include at this time other behavioral aspects of reproduction.
The presentation that follows attempts to establish the impact
and causes of sexual dysfunction in humans, to discuss methods of
assessing human sexual behavior, and to indicate the difficulties
associated with investigations of human sexuality when there are
neither controls nor standard norms. The evaluation of sexual
behavior patterns in animals are presented as simple tallies of specific
motor responses; however, it will be emphasized that many elements
of human sexual behavior are unique, having no animal counterpart,
thus making uncertain any extrapolation of data on animal sexual
behavior to humans.
-------
124
Definition and scope. The study of sexual behavior encompasses
the measurement of normal and abnormal function as well as the
identification of the factor(s) responsible for the impairment of
sexual behavior. The quantitative measurement of sexual functioning
involves the establishment of norms or averages for groups and for
the individual and most often focuses on coitus itself. For humans
more extensive and varied measures are necessarily employed, which
include sexual imagery, sexual fantasy, varieties of overt sexual
experience, self concept and gender identity, assessment of inter-
personal relationships, choice of sexual object, and level of sexual
skill.
For the most part, animal models available today do not provide
data that might be required for assessing human sexual functioning
and for identifying the factors responsible for imparied expression of
sexual behavior. Nevertheless, important advances made in the study
of animal sexual behavior provide at least for the initial screening of
toxicants that could deleteriously affect human sexual conduct. To
identify the factors responsible for normal and impaired sexual
behavior, investigators of animal sexual behavior have identified
three separate components: sexual attractiveness, sexual initiative,
and sexual responsiveness. The current working assumption is that
these factors have a much broader generalizability across species than
any isolated species-typical behavioral response or activity (e.g.,
mounting). The evidence and arguments favoring this working
assumption have been set forth persuasively by Beach (24), who uses
the alternative terminology of attractivity, proceptivity, and recep-
tivity to designate the three factors. Accepted systems of measure-
ment have been worked out for a variety of laboratory animals
including the rat and macaque monkeys (25—27).
It should be realized from the outset that manifest sexual
behavior reflects the functional integrity of a broad system com-
prising elements of drive and reward, perception, sensory function,
motor performance, the physicochemical actions of gonadal hor-
mones on neural and somatic tissues, and, finally, central nervous
system processing and coordinating of the interactions of all of these
elements. An efficient screening system utilized for detection of
toxic effects should attempt at some stage to distinguish between
general motoric disability (ataxia) and impairment of specific sexual
reflexes; between general lassitude and loss of specific sexual interest
or motivation; and between the impaired production of gonadal
hormones and the impaired actions of these hormones. At this tune,
however, no simple and efficient test of sexual behavior or tests
-------
125
designed to measure sexual attractiveness, initiative, and respon-
siveness automatically determine whether elements of the broad
system are impaired either as a result of general debilitation or
selectively and specifically "with regard to sexual performance.
Currently available tests, while not permitting decisions about the
specificity of the effect of a toxic substance, can serve as early
warning signals that normal reproductive function has been impaired.
Impact of sexual dysfunction on humans. Data on the incidence
of human sexual dysfunction and its spontaneous remission are
neither extensive nor very reliable. Certainly the most common
clinical problems are primary and secondary anorgasmia in females
and ejaculatio praecox and erectile impotence in males. These
disorders, however, represent extremes of dysfunction that are
unacceptable to most humans, and those so affected commonly seek
clinical help. Less extreme forms of inadequate sexual response
clearly exist and are often tolerated, but only in the sense that
professional counseling is not sought. Even though many individuals
are reluctant to seek professional help, the importance of sexual
behavior and sexual gratification to the overall quality of life and
individual well-being is generally recognized. Many individuals are
willing to relinquish reproductive capabilities (through vasectomy or
other contraceptive means) to enjoy fewer restrictions on sexual
activity. Few people, however, will relinquish sexual gratification to
gain contraception.
The importance attached to sexual gratification by individuals in
our society implies that sexual inadequacy, even when tolerated, may
not be without serious consequences. Our monogamous social system
depends in a very fundamental way upon a sexual contract between
two individuals. Failure to achieve, or even reduction of, sexual
satisfaction seriously threatens the interpersonal relationship, as the
growing number of marriage counselors recognize. How much of the
growing sexual and marital discord is attributable to sociopsycho-
logical factors and how much might be attributable to disturbances
in the physiological systems underlying sexual performance is not
known. The possibility exists, however, for toxic substances in the
environment to cause disturbances in sexual performance and
thereby contribute to interpersonal discord.
Causes of impaired sexual performance. Sexual dysfunction in
human beings is poorly understood. Many sexual disorders are
primarily psychogenic and respond well to psychological treatment.
Others that are resistant to psychological approaches seem to
originate in specific genetic factors, early experience, or a combina-
tion of genetic and experiential determinants.
-------
126
Endocrinopathies, especially abnormalities of the gonadal hor-
mones, have marked influences on the pattern of sexual behavior
displayed by animals, and although their influence is less well
described for humans, it cannot be said that their role is negligible.
For both the male and the female, inadequate gonadal hormone
activity commonly results in deficient sexual performance. Sub-
normal effectiveness of gonadal hormones can be due to (a)
deficiencies in production, (b) deficiencies in bioavailability, and (c)
deficiencies in target organ sensitivity and/or responsiveness.
Androgen deficiency affects male behavior in two distinct ways.
First, during early stages of development (probably before birth in
humans) deficiencies in androgen lead to incomplete development of
central neural and peripheral somatic structures essential for the
expression of masculinity and male behaviors including, but not
limited to, male sexual behavior. Second, during adolescence and
adulthood, deficiencies in androgen are associated with reduced
sexual responsiveness and sexual initiative.
Behavioral disorders associated with excessive androgen have not
been identified for the male, although there are recurrent suggestions
that excessive amounts during early stages of development lead to
permanent androgen insensitivity. In the female, however, excessive
androgen during early developmental stages leads to the development
of masculine behavioral and somatic characteristics and, in some
species, to the suppression or loss of feminine behavior traits. This
suppression of feminine traits can include sexual behavior, and
female sexual responsiveness can be only reduced. In adulthood,
excessive androgen in female humans may lead to measurable
somatic virilization (such as hirsutism and clitoromegaly) without
any marked effect on psychological and behavioral traits. Increases in
sexual initiative and responsiveness may be large enough to be
distressing and disruptive to an established interpersonal relationship.
In nonhuman primates, androgens have also been implicated in the
control of proceptivity (sexual initiative), but not in the control of
receptivity. In other mammalian animals, excessive androgen in
adulthood may cause a sharp increase in the frequency of malelike
mounting activity and aggression with or without concomitant
alterations in female sexual behavior. Most of the psychological and
somatic changes induced by excessive androgen in adulthood are
partially or totally reversible when the hormonal excess is eliminated;
however, some of the more dramatic somatic changes (e.g., voice
changes and hirsutism) are irreversible.
Estrogen and progestagens play essential but incompletely
understood roles in the regulation of female sexual response. These
-------
127
hormones are produced and secreted in much larger amounts by the
ovaries than by the testes or the adrenals in normal physiological
conditions. Whereas human female sexual behavior does not depend
entirely upon the actions of estrogens and progestagens, its expres-
sion is greatly facilitated by their actions on both central and
peripheral neural and somatic tissues. In animals, especially the
common laboratory forms, the ovarian hormones are much more
essential to the expression of female sexual behavior than in human
beings. Generally, the effective estrogen is estradiol and the effective
progestagen is progesterone. These two steroid hormones act
synergistically in the induction of both proceptivity and receptivity
in rats, mice, hamsters, and guinea pigs. The two hormones may also
act antagonistically, however, and which relationship obtains de-
pends upon whether or not the estrogen has been free to act for a
specifiable period of time without any concurrent actions of a
progestagen. The synergistic relationship depends upon the sequen-
tial action of an estrogen followed (usually 36 to 48 hours later) by
the action of a progestagen. An antagonistic relationship will be
evidenced whenever a progestagen and an estrogen are both present
throughout the period of observation or study.
Excessive estrogenization acts to lengthen the period or duration
of receptivity and proceptivity. An established norm of eight hours
for the duration of receptivity in a colony of rats can be extended to
12 or 14 hours by excessive estrogenization. In extreme cases,
excessive estrogenization can extend receptivity indefinitely.
Excessive progesterone has no measurable effect if the period of
stimulation is brief. If the period of excessive stimulation is
protracted, however, receptivity and proceptivity can be indefinitely
suppressed or inhibited. The antagonistic effect of progestagens is
transitory and reversible when these hormones are brought back to
physiologic concentrations.
It is difficult but not impossible to distinguish between the
antagonistic effects of excessive progestagen and a deficiency in
estrogenization. A deficiency of estrogen, like excessive pro-
gestagenization, has the primary characteristic of weak or absent
female sexual response. A distinction between the two possible
causes of impaired sexual response can be made by institution of
appropriate experimental hormone administration to ovariectomized
females.
Excessive estrogen or progestagen during early periods of
development can produce permanent deficiencies in female sexual
response in a variety of laboratory animals. Comparable data for
human beings do not exist.
-------
128
Sequelae of estrogen and/or progestagen excess or deficiency in
males have not been well worked out. Supraphysiologic levels of
estrogen have been administered to human males in cases of prostatic
cancer. Sexual drive and erectile potency sometimes decline in these
cases, presumably because the estrogens block the release by the
pituitary of testis-stimulating hormone, and an androgen deficiency
results. Similar effects could be obtained in some laboratory animals
(the guinea pig), but not in others (the rat, in which excessive
estrogens mimic androgens in the potentiation of male sexual
activity).
The effects of hormone excess and deficiency could occur when
chemical substances mimic or antagonize physiological actions of the
relevant gonadal hormone. Other chemical agents could enhance the
degradation of steroidal hormones in the liver or kidneys and thereby
reduce or limit their effectiveness. Still other chemicals could either
act on the hypothalamic-pituitary system to modify the release of
trophic substances essential for the normal production of the gonadal
hormones or act directly upon the glandular tissues responsible for
their production.
Many factors, aside from alteration of or interference with
hormonal support, can act to impair sexual behavior. These factors
are difficult to assess in standard laboratory tests, either because no
suitable animal model can be found or because appropriate assess-
ment would involve procedures too elaborate and costly for
routinization. Although testing for alteration or interference with the
hormonal support of sexual behavior assesses only a limited aspect of
requirements for human sexual adequacy, it has the advantage that
the information gained is reliable, quantitative, and amenable to use
in estimating the risk to human sexuality posed by specific chemical
substances.
Qualitative evaluation of risk potential
Interspecies comparisons. Requirements for genetic variability in
the test model animal that approximates that encountered in the
human population have already been discussed in the section titled
"Interspecies Comparisons" in Chapter 4. Highly inbred strains ought
to be generally avoided unless several are used to determine the range
of sensitivity to the test substance.
Known and suspected differences among species in the manner in
which gonadal hormones regulate sexual behavior mandate the use of
more than one species. For example, the major androgen produced
by the testis is testosterone in most mammals. This hormone is
-------
129
the normal male rat,
or via conversion to
mediating male sexual
metabolized within somatic and neural cells to a variety of other
steroids including estradiol and dihydrotestosterone. In some species,
like the rat, the estradiol derived from bioconversion of testosterone
is a potent stimulator of male sexual behavior in the adult and a
potent masculinizer of the developing brain in the fetus and neonate.
In contrast, in the guinea pig this estrogen metabolite of testosterone
is without any measurable stimulating effect on male sexual behavior
when it is given to castrated adults. The view is widely held that the
display of male sexual behavior depends upon the intracellular
conversion of testosterone to estradiol in
whereas testosterone acts either directly
dihydrotestosterone on the neural tissues
behavior in the guinea pig.
The "rat model" for cellular utilization of testosterone (by
conversion to an estrogen) is valid for hamsters and some but not all
inbred strains of mice. The "guinea pig model" is valid, based on very
limited data, for the rhesus monkey and also for humans.
This species difference is important because erroneous conclu-
sions are possible if testing is limited to a single species. Any putative
toxicant that blocks intracellular aromatizing enzymes needed for
bioconversion of testosterone to estrogen would impair adult male
rat sexual behavior, but the same compound would not likely have
an effect on sexual behavior of male guinea pigs, rhesus monkeys, or
humans.
Other species differences, too numerous to detail here, include
differences in the role of specific neural structures, in the contribu-
tion of specific neurotransmitters, in the amount and kind of carrier
protein that is present in the bloodstream and binds and transports
the steroid hormones, in the chemical structure of the pituitary
trophic hormones, and certainly in the form and normal frequency
of sexual expression. All of these species differences argue for the use
of more than one species in screening for toxicity as well as for
judicious choices when only a few are to be used. In short, the choice
of a species to use as a model animal places profound and subtle
limits on evaluating the toxic consequences of any chemical agent,
and these limits have to be reckoned with.
Certain relatively simple and easy-to-conduct tests could serve as
a preliminary screen to indicate the degree of likelihood of an agent's
affecting either the early sexual differentiation or adult expression of
sexual behavior. Based on the assumption that chemical substances
that pass the placenta and gain access to the fetal tissues are more
likely to affect early development than those that fail to pass
-------
130
through the placenta, a relatively simple and efficient study of the
distribution of the radioactively labeled chemical substance could be
carried out. It is also reasonable to use radiolabeled material in adult
animals to determine whether the substance crosses the blood-brain
barrier and has the potential of acting directly on nervous tissue.
These simple tests, of course, are not specific indicators that sexual
behavior would be altered by the putative toxicant. Positive findings
from these tests would merely serve the purpose of indicating
increased likelihood.
Other considerations. It is reasonable to assume that a wide
variety of other factors are important in facilitating extrapolation of
animal tests of a toxicant to the estimation of risk to humans. These
include dosages of putative toxicant used, route of administration,
duration and frequency of exposure, species thresholds and sensitivi-
ties to the chemical substance, and specific pharmacokinetics and
pharmacodynamics of the test compound. Whenever information
exists for humans on any of these factors, either for the specific test
compound or a closely related substance, an effort should be made
to select an animal model that most closely parallels the human to
increase the applicability of the animal test results.
Despite the advantages of objectivity, ease of administration,
reliability, and wealth of background information, tests of animal
sexual behavior in the present context have severe limitations. First
and foremost is the high degree of uncertainty that results of animal
tests could be extrapolated to human sexual behavior. It is likely that
extrapolation would be good if a putative toxicant completely
blocked the neurological actions of the sex hormones (especially the
androgens), since hormonal support for sexual behavior and for the
fetal differentiation of sexual and/or sex-related behavior is a factor
common to both animals and humans. However, humans and animals
differ greatly in the numbers and kinds of nonhormonal factors
influencing the expression of sexual behavior. Accordingly, when a
putative toxicant acts only on one or a subset of nonhormonal
factors, there is a strong likelihood that animal test results will not
correspond to effects (or lack thereof) on human sexual expression.
A second area of concern is the nearly total lack of background
information on effects of known toxicants on sexual behavior in
either animals or humans. This deficiency thwarts any present
attempt at formulation of procedures for quantitative risk assessment
based on findings from animal tests. This situation can be remedied
only by providing encouragement of the appropriate research on
animal models as well as intensive studies of humans exposed to
known toxic agents.
-------
131
Animal studies
Evaluation of sexual behavior in adulthood. Observations of
sexual behavior in adult animals can be made by easily trained
nonexperts. Useful assessments of the status of sexual behavior can
be made from simple tallies of the frequency of occurrence of
specific motor responses and the latent period from the beginning of
a standardized test to the occurrence of the specific response. These
are the operational measures of initiation, attractiveness, and
responsiveness.
The procedures described in this section are designed to permit
reliable, sensitive analyses of the effects of potentially active
chemical substances on male sexual behavior. A considerable body of
knowledge gathered in the last 60 years reveal's that the sexual
patterns of rats and guinea pigs can provide such data. Further, the
existence of extensive data bases on these two species provides the
possibility of a preliminary indication of mechanism of action
underlying observed treatment effects, since determinants of various
aspects of these complex patterns have received much study. Should
more extensive and expensive testing of a chemical be indicated,
dogs, nonhuman primates, or other species may be appropriate.
Methods described below can be adapted to such species using
behavioral testing procedures described by Dewsbury (28) and in the
references therein.
In all tests, one sex should be treated so that treatment effects
may be detected uncomplicated by effects of the agent's acting on
the opposite sex. Where the probability of a treatment effect is quite
low, it may be more economical to combine procedures for male and
female treatments into a single protocol. However, the risk that three
rather than two studies may be required if such procedures are used
should be recognized.
Assessing sexual behavioral patterns of males. Effects of various
toxicants can be evaluated as they alter the normal complex
behavioral patterns of male rats and guinea pigs. There are some
considerable advantages to toxicological inquiry in studying the
behavior of male animals that have been castrated and given
physiological hormone replacement and exogenous testosterone. This
procedure obviates the possibility that impaired sexual performance
might be due to toxic insult to the hypothalamic-pituitary-testicular
axis or to the testis itself. However, inasmuch as castration and
replacement therapy are complicated techniques in themselves,
testing the intact animal should serve as an adequate preliminary
screen.
-------
132
In studying the normal copulatory behavior of laboratory rats,
three classes of events are generally distinguished — mounts,
intromissions, and ejaculations. With the first the male mounts the
female from behind, displays shallow pelvic thrusting, but neither
gains vaginal penetration nor displays the stereotyped pattern of
dismounting. Intromissions begin similarly, but the male achieves a
single deep thrust and dismounts in a vigorous and stereotypical
pattern. Ejaculations occur only after several intromissions and are
characterized by an intravaginal thrust that is longer and deeper than
that of intromissions. Sperm are transferred only on ejaculations.
The male mounts the female during mounts, intromissions, and
ejaculations, but the three classes of events are distinguished as just
indicated. During pair mating copulatory events occur in "ejacu-
latory series," with each series terminated by an ejaculation and
separated from a resumption of copulation by a postejaculatory
refractory period. In standard testing cages, males normally display a
mean of approximately seven ejaculatory series before attaining an
arbitrary, but standard, satiety criterion of 30 minutes with no
intromissions or ejaculations.
Standard measures of male copulatory behavior include mount
latency (ML), time from start of a test to the first mount or
intromission; intromission latency (IL), time from the start of a test
to the first intromission; ejaculation latency (EL), time from the first
intromission of a series to its terminal ejaculation; intromission
frequency (IF), number of intromissions in a series; mount frequency
(MF), number of mounts in a series; mean interintromission interval
(Mill), mean interval separating the intromissions within a series; and
postejaculatory interval (PEI), time from ejaculation to the next
intromission. Male receptivity may be quantified by dividing the
number of male chase and follow-bouts by the total number of
female approaches (29).
Because various of these measures can be affected selectively,
specifically, and in combination, an accurate interpretation of a
treatment effect requires a full complement of these measures. As an
example, suppose a treatment interfered with the process of penile
erection. Males with such problems often mount females at rates
much higher than normal as they repeatedly attempt to effect
intromission. Without a full complement of measures, such an effect
might be mistaken for an increase in libido rather than as a deficit.
Similarly, a treatment that alters Mill may secondarily affect IF. In
addition, with a full set of measures, one can evaluate the control
group in relation to animals used in previous studies (see references
-------
133
below) to ensure that it is providing an appropriate baseline for
comparison.
Full descriptions of copulatory behavior in male rats can be
obtained in Beach and Jordan (30), Dewsbury (27), Larsson (31),
and Sachs and Barfield (32).
Copulatory behavior in male guinea pigs differs from that in male
rats in several important respects. First, whereas rats display but a
single intravaginal thrust on each mount with intromission, guinea
pigs display repetitive thrusts on a single insertion. Second, whereas
male rats rarely, if ever, ejaculate on the first mount with
intromission, such occurrences are more frequent in guinea pigs.
Third, although male rats normally display several ejaculations per
.test session, the occurrence of the first ejaculation generally
effectively terminates copulatory activity in guinea pigs. In other
respects, the copulatory patterns of male guinea pigs are quite similar
to those of male rats. Similar measures can be used.
Descriptions of copulatory behavior in guinea pigs can be found
in Young (33) and Young and Grunt (34). Various measures of
preliminary aspects of courtship and mating described in these papers
may be useful.
Test of copulatory behavior should be conducted during the dark
phase of the diurnal cycle. By testing during the second half of the
dark phase, behavior generally is more reliable, quicker, and less
variable — making for a more efficient and sensitive test (35). Tests
should be conducted at approximately the same time on each day.
Males and females should be familiar with the testing arenas via
introduction several times on days before test days.
In tests for male behavior, males are generally placed in the
arenas for five to ten minutes, after which the female is introduced,
effectively beginning the test. Tests may be terminated and scored as
negative if there is no copulatory activity within a predetermined
time (e.g., 15 minutes).
Tests of guinea pigs should be terminated at the occurrence of
ejaculation. Those of rats should be continued for two or three
ejaculatory series. Such tests may require an average of 45 minutes in
rats. It may be feasible to test several pairs of rats simultaneously in
cages close to each other, if the only behavioral patterns to be scored
are those discussed above.
For reasons of reliability and predictability, it is recommended
that female mating partners be brought into behavioral estrus with
exogenous hormones. Female guinea pigs must first be ovariecto-
mized; this may or may not be done with female rats. For either
-------
134
intact female rats or spayed guinea pigs, good results can be obtained
with an intramuscular injection of 0.1 mg of estradiol benzoate three
days before testing and 1 mg of progesterone approximately six
hours before testing. Somewhat lower doses can also be used.
Females should be placed briefly with a vigorous, nonexperimental
"indicator" male immediately before testing to ensure that the
injection regimen has been effective in inducing receptivity. A single
female rat in estrus can be used to evaluate sexual performance of at
least three males. A single female guinea pig should not be used for
more than two males.
There are many factors, both quite specific and highly non-
specific, that' can alter copulatory behavior. If there are gross
increases or decreases in body weight or activity levels, changes in
sexual behavior are probably secondary to more general effects. If
body weight and general activity is near normal and copulatory
behavior is altered, however, greater specificity of action probably is
indicated. Some indication of the nature and degree of effect can be
determined by considering the constellation of measures altered, the
magnitude of effect, and reversibility. By comparing these changes to
those described in the literature as resulting from other treatment,
some preliminary indication as to probable mechanism of action can
be gained. Any alteration requires some further analysis. Such
subsequent studies may be directed at analyzing neural, endocrine,
and other systems to determine whether or not the effect seems
appreciable and likely to affect humans.
Assessing sexual behavior patterns of females. A substantial and
useful background of behavioral data exists for both rats and guinea
pigs from a number of inbred strains as well as genetically
heterogenous stocks. It is possible to evaluate proceptivity, recep-
tivity, and attractiveness in female rats in a single test paradigm with
a stud male and to evaluate receptivity in the female guinea pig.
The intact and normally functioning female rat displays a period
of estrus ("heat") that lasts from 6 to 11 hours about every 4 to 5
days. As long as the female is not mated, estrus recurs regularly.
Recurrent estrus also occurs in the unmated female guinea pig, but
the interval between receptive episodes lasts 14 to 17 days. In both
species, sexual behavior depends upon appropriate ovarian secretion
of estradiol and progesterone. When the ovaries are removed, sexual
behavior is no longer displayed. Preceptive and receptive behaviors
are displayed in close temporal proximity and have similar hormonal
requirements (36).
-------
135
The behavioral response indicative of normal receptivity is the
lordosis posture assumed by the female during mounting by a male
partner. Degree of receptivity is estimated in a quantitative fashion
by dividing the number of lordosis responses displayed by the female
by the number of times she is mounted by her male partner in a
standardized test. This derived measure is called the receptivity
quotient or, alternatively, the lordosis quotient. The measure is more
useful in the rat than in the guinea pig, because male rats are
normally multiple mounters, whereas male guinea pigs often mount
only once during a mating test. For the female guinea pig, therefore,
an alternative procedure for quantifying receptivity is often used.
The procedure, described fully elsewhere (37, 38), involves manual
stimulation of the animal's rump and perineum by the human
observer and measurement of the degree or duration of the lordosis
response to such stimulation. During mating with a stud male, female
attractiveness may be quantified by measuring the latency between
introduction of the female and a male approach, follow, and mount
(29).
Proceptivity can be measured quantitatively in the female rat by
recording the frequency and timing of displays of a variety of motor
patterns including female solicitations and approaches to the male
partner, darting, hopping, and ear vibration. These preceptive
patterns generally are displayed just prior to the occurrence of male
mounting responses, but they may occur at any time when the male
is quiet or inactive.
Full descriptions of female rodent sexual behavior can be found
in Diakow and Dewsbury (39); McClintock and Adler (25);
McClintock, Ansiko, and Adler (40); and Madlafousek and Hlinak
(26).
Evaluation of sexual responses in the intact female requires
constant and frequent monitoring of individual animals. This is
essential because the occurrence of the behavior is restricted to a
short and specific period of the ovarian cycle. The behavior normally
is expressed only during the time the follicle is undergoing its final
preovulatory swelling. The ovarian cycle is usually monitored by
taking daily vaginal smears for cytological evaluation, and sexual
behavior usually is displayed during the proestrous smear or the
transition between proestrous and estrous smears.
The procedure of monitoring the ovarian cycle by daily vaginal
smears is cumbersome, time-consuming, and not very precise with
respect to the assessment of sexual behavior. When individual females
are tested for receptivity and proceptivity at an arbitrary time
-------
136
relative to a particular vaginal cytology, some may be at the
beginning of the period of receptivity, some in the middle, and some
near the end. Others may not yet have reached the receptive stage,
and for those in various segments of the period the quality of
receptive behavior will vary accordingly. Furthermore, in the intact
female, impairment or absence of sexual response could be due to
impairment of pituitary gonadotrophic activity, disordered ovarian
production of steroids, or impairment of the response of relevant
neural centers to the gonadal hormones.
Undesirable variability as well as uncertainty about the cause of
impaired sexual response can be reduced by assessing sexual behavior
in spayed females suitably treated with injections of estradiol and
progesterone. Usually females are brought into good states of
receptivity by a single subcutaneous injection of estradiol benzoate
followed 48 hours later with an injection of progesterone. All
animals to be tested can then be evaluated at an exact time (usually
six hours) after the progesterone injection.
The artificial induction of sexual responses has to be done with
precision and with concern for hormonal stimulation that closely
approximates the normal physiological pattern. Administration of
excessive amounts of estrogen and progesterone could mask or
override subtle derangements induced by a toxic substance. If spayed
animals are used for assessment of sexual behavior, great care must
be exercised to ensure that physiological doses of estrogen and
progestagen are administered. Reference to the literature on experi-
mental analysis of female rodent sexual behavior will not be helpful
as a guide to proper hormone treatment, since suprathreshold dosage
regimens are usually used, and these are not appropriate for screening
toxicants. In any attempt to identify damaging actions of a putative
toxicant, the investigator should be cautious about exceeding 1 fig of
estradiol benzoate and 0.1 mg of progesterone per adult animal. The
best general rule to follow is to conduct an initial parametric study
on the specific breed or strain to be used and to determine the
minimum hormonal requirements for induction of estrous behaviors
in a specified percent of the population.
Assessment of human sexual behavior: surveillance and
epidemiological studies
Preliminary comments. Direct assessment of human behavior is
essential for evaluating the behavioral effects of environmental
toxicants. The extrapolation of animal studies to human behavior is
limited for a variety of reasons, (a) Many aspects of human sexuality
-------
137
and reproductive behavior are unique and have no obvious animal
counterpart (41, 42). (b) While compounds such as steroids do affect
sexual motivation in both animals and humans, their behavioral
manifestations in humans are often quite different from their
manifestations in animals, (c) Human behavior may be disrupted at
lower toxicant levels than would be expected from animal studies.
(d) The exposure of the general population, but especially workers,
to the compound may be greater in fact than originally estimated
(also see "Other Considerations").
We present several different methods for assessing the effect of a
toxicant on human sexual behavior. Because direct controls may not
be possible for practical or ethical reasons, each method has its own
weakness. Therefore, we have proposed a variety of methods and
suggest that they be used concurrently if at all possible. This extra
effort would be justified particularly when the potential benefits of a
compound are high, but also when animal toxicological screening or
analysis of the compound's structure indicates that the potential risk
to human behavior may also be high. In any case, the behavioral
assessment procedures for humans need not be cumbersome and can
be incorporated in any procedure or physical exam designed to
monitor the effects of a putative toxicant on reproductive function.
Behavioral surveillance of humans potentially exposed to a
reproductive toxicant. Ideally, new compounds would be released
and used at first on a limited basis. Then, changes in sexual
satisfaction and function could be assessed prospectively with
adequate controls. The sexual experience of the exposed group,
perhaps production workers who would be exposed to higher
concentrations, could be compared with a matched group of similar
workers in an area or plant where the compound was not yet in use.
This comparison should be made between two groups of workers in
the same plant or location. If this is not possible, the two groups
should be matched for factors known to correlate with sexual
attitudes and behavior such as socioeconomic status, cohort, eth-
nicity, religion, and environmental factors. (Industry should use an
epidemiologic consultant to determine the matching criteria, sample
size, and duration of surveillance appropriate for the amount of
natural variance in the proposed measures of sexual behavior.)
If limited release is not warranted ethically or practically and a
general release occurs, it would be necessary to monitor sexual
satisfaction and the incidence of dysfunction before as well as after
the compound is released. The large population variability in normal
sexual behavior may make this procedure more sensitive than a
-------
138
comparison between groups. Furthermore, as the behavior of each
person is compared to his own normal pattern, it may be possible to
identify a subpopulation of particularly sensitive individuals.
These control procedures are essential for evaluating the effect of
a toxicant on human sexual behavior, because standard norms are
not currently available as a basis for comparison as they are for such
physical variables as sperm count or menstrual cycle length. As the
number of controlled studies increases, it may be possible to use the
data from control groups to develop normative statistics for future
evaluations.
The frequency of sexual intercourse is not a good indicator of
sexual satisfaction by itself; it is also necessary to evaluate sexual
arousal, initiation, and changes in erotic imagery and to identify
specific sources of sexual dysfunction. For example, there was little
agreement about the nature of changes in women's sexual motivation
over the menstrual cycle until studies focused on behavior of the
woman herself and her sexual initiation and fantasy rather than on
the frequency of intercourse (43, 44).
Either an interview or a short questionnaire can assess sexual
satisfaction and function. It is important that the interviewer be
trained in interview techniques for sexual counseling. Short courses
are currently available for medical and lay personnel in most
academic medical centers (Marriage Counseling Center of the
University of Pennsylvania has a list). Alternatively, there are a
variety of short questionnaires that correlate well with such
physiological measures of sexual arousability as penile tumescence
and vaginal lubrication (45, 46) and that generate a similar profile
whether completion of the questionnaire is mandatory or voluntary
(47, 48).
Another approach to the assessment problem is based on
epidemiological data. The incidence of cases involving sexual
dysfunction reported to such institutions as mental health clinics,
local physicians, or plant infirmaries can be recorded and used as a
normative data base. This baseline could be compared with the
frequency of reported cases following the release of a new
compound. Again, an epidemiologic consultant should determine
whether the sample from available institutions would be large enough
to detect a toxic effect.
Evaluation of human sexual behavior following exposure to a
known toxicant. Many compounds have been established as physio-
logical toxicants but have not yet been assessed for behavioral effects
in humans. Estrogenic compounds such as DBS and DDT may affect
-------
139
the sexuality of women, while organopesticides that are neurotoxins,
such as carbaryl, an acetylcholine esterase inhibitor, may affect male
erectile function.
If a population has been exposed to such compounds or is
suspected of being at reproductive risk, behavioral assessment can be
made at the time that a physiological assessment is being made. The
same personnel could do this, provided that they have been trained
in interview techniques. Behavioral assessment under these ex post
facto conditions is particularly difficult because knowledge of
exposure to a toxicant can distort the retrospective assessment of
sexual satisfaction and behavior. Therefore, trained personnel, an
evaluation immediately after the exposure, and established norms
would each help to reduce this bias. In any event, an unexposed
control population should be evaluated using the identical retrospec-
tive procedures and matched to the target population for such
variables as socioeconomic status, ethnic group, and local environ-
ment.
Risk assessment. If any significant alterations are found, expo-
sure to the toxicant should be discontinued to assess the reversibility
of the effects. Furthermore, the mechanism of action will need to be
identified to evaluate a risk/benefit ratio. For example, it is possible
that erectile function could be impaired through a direct impairment
of cholinergic mechanisms or indirectly through an increase in
depression or sense of fatigue (49). Nocturnal penile tumescence
would aid in a differential diagnosis, as erectile function during sleep
is not impaired by psychogenic factors. Human sexuality is particu-
larly sensitive to disruption by many environmental and psycho-
logical factors that are not specifically sexual themselves; most
instances of sexual dysfunction encountered in the clinic are not the
result of a direct organic cause. Therefore, the mechanisms of any
impairment of sexual performance or satisfaction will have to be
determined before a risk/benefit ratio can be assessed.
Priority areas for future research
Few experimental studies have been made of the effects of toxic
substances on the sexual behavior of laboratory animals. A few
recent references (50-61) are included in the listing at the end of
this chapter, but they deal primarily with effects of drugs like
cannabis, alcohol, and morphine. A search of the literature between
1978 and 1980 revealed only two references dealing with other
agents, one on cyanogenic substances (62) and the other on lindane
(63). In addition, no systematic evaluations of sexual behavior have
-------
140
been conducted on humans known to have been exposed to toxic
substances either in adulthood or prenatally. This unfortunate
circumstance, that parallel studies have not been carried out on
intentionally exposed animal subjects and on accidentally exposed
human beings, severely limits the capability to formulate either
qualitative or quantitative risk assessments for sexual functioning.
Basic research on human sexual behavior should be strongly
encouraged at this time so that appropriate demographic norms and
standards can be established. Adequate information on these matters
has not been developed despite the pioneering efforts of Kinsey in
the late forties. In addition, changes in concepts, data gathering
techniques, and attitudes require modernization of the data base. As
pointed out in earlier sections of this discussion, neither measure-
ment of number of offspring produced, frequency of coitus, or even
frequency of orgasm are adequate as indicators of human sexual
functioning. There is a strong need to develop epidemiological
studies of human sexual behavior in its broadest scope and in terms
most meaningful to human welfare and to the stability of interper-
sonal relationships.
The scope of investigations of animal sexual behavior should be
broadened. Efforts to establish models permitting better measure-
ment of sexual attractiveness, sexual motivation or desire, and even
sexual gratification should be encouraged. Moreover, sound para-
metric data on the effects of known environmental toxicants ought
to be vigorously pursued. These studies could be carried out
profitably at this time even with the limited number of behavioral
measures currently available, and there should be strong support for
such studies on a variety of species. The limitation of data bases, no
matter how extensive, to rats and guinea pigs poses a serious obstacle
to flexible choice of alternative models that may in fact be more
appropriate to human problems.
Finally, even from the relatively limited standpoint of sexual
behavior, more information is needed on how different classes of
chemical substances interact with central neural tissues on the
cellular levels. Information of this sort is fundamental not only to
interpretation of toxicant effects on behavior, but also to sound
hypothesis formulation and to development of a framework that
would permit prediction of the likely biological effects of a putative
toxicant.
-------
141
REFERENCES
1. Rausch, L., Klein, J., Shaffer, S. Q. and Stein, Z.: Effect of chlorination of
drinking water on perinatal'mortality in New York villages, 1968-1977.
Abstract. Am. J.Epidemiol. 112: 439, 1980.
2. Levine, R. J., Symons, M. J., Balogh, S. A., Arndt, D. M., Kaswandik, N. 'T.
and Gentile, J.W.: A method for monitoring the fertility of workers: I.
Method and pilot studies. J. Occup. Med. 22: 781-791, 1980.
3. Levine, R. J., Symons, M. J., Balogh, S. A., Milby, T. H. and Whorton,
M. D.: A method for monitoring the fertility of workers. II. Validation of
the method among workers exposed to dibromochloropropane. J. Occup.
Med. 23: 183-188, 1981.
4. Slone, D., Shapiro, S. and Miettinen, O. S.: Case-control surveillance of
serious illnesses attributable to ambulatory drug use. In: Epidemiological
Evaluation of Drugs, F.Colombo, S.Shapiro, D.Slone, and G.Tognoni,
Eds., Elsevier North-Holland Biomedical Press: Amsterdam; pp. 59—82,
1977.
5. Rosenberg, L., Shapiro, S., Slone, D., Kaufman, D. W., Miettinen, O. S. and
Stolley, P.O.: Thiazides and acute cholecystits. N. Engl. J. Med. 303:
546-548, 1980.
6. Mitchell, A. A., Rosenberg, L., Shapiro, S. and Slone, D.: Birth
defects related to use of Bendictine in pregnancy. I. Oral clefts and
cardiac defects. J. Am. Med. Assoc. 245: 2311-2314, 1981.
7. Selikoff, I. J., Churg, J. and Hammond, E. C.: Relation between exposure to
asbestos and mesothelioma. New Engl. J. Med. 272: 560-565, 1965.
8. Lee, A. M. and Fraumeni, J. F., Jr.: Arsenic and respiratory cancer in man:
An occupational study. J. Natl. Cancer Inst. 42: 1045-1052, 1969.
9. Harlap, S,, Shiono, P., Ramcharan, S., Berendes, H. and Pellegrin, F.: A
prospective study of spontaneous fetal losses after induced abortions. New
Engl. J. Med. 301: 677-681, 1979.
lO.Notari, R. E.: Biopharmaceutics and Pharmacokinetics: An Introduction.
3rd Edition, Marcel Dekker: New York; 329 pp., 1980.
11. Wagner, J. G.: Biopharmaceutics and relevant pharmacokinetics. Drug
Intelligence Publications: Washington, D.C.; 375 pp., 1971.
12. Gibaldi, M. and Perrier, D.: Pharmacokinetics. Marcel Dekker: New York;
325 pp., 1975.
13. Bischoff, K. B.: Some fundamental considerations of the application of
pharmacokinetics to cancer chemotherapy. Cancer Chemother. Rep. Part 1
59: 777-796, 1975.
14. Himmelstein, K. J. and Lutz, R. J.: A review of the applications of
physiologically based pharmacokinetic modeling. J. Pharmacokinet.
Biopharm. 7: 127-145, 1979.
IS.Dedrick, R. L.: Animal scale-up. J. Pharmacokinet. Biopharm. 1: 435—461,
1973.
16. Dedrick, R. L. and Bischoff, K. B.: Species similarities in pharmacokinetics.
Fed.Proc. 39: 54-59, 1980.
-------
142
17. Gehring, P. J., Watanabe, P. G. and Blau, G.E.: Pharmacokinetic studies in
evaluation of the toxicological and environmental hazard of chemicals: In:
Advances in Modern Toxicology: New Concepts in Safety Evaluation,
M. Mehlman, H. Blumenthal, and R. Shapiro, Eds., Hemisphere Publishing
Corporation: New York; pp. 195-290, 1976.
18. World Health Organization. Chemobiokinetics and metabolism. In: Environ-
mental Health Criteria 6. Principles and Methods for Evaluating the
Toxicity of Chemicals. Geneva; pp. 116-177, 1978.
19.Filov, V. A., Golubev, A. A., Liublina, E.I. and Tolokontsev, N.A.:
Quantitative Toxicology. John Wiley: New York; 462 pp., 1980.
20. Young, J. F. and Holson, J. F.: Utility of pharmacokinetics in designing
toxicological protocols and improving interspecies extrapolation. J.
Environ. Pathol. Toxicol. 2: 169-186, 1978.
21.Souchay, A. M., Rice, P. C., Rourke, I.E. and Nesbitt, R.E. L.: Glucose
metabolism and transfer in the human placenta. In: Chemical Engineering
in Medicine, D.D. Reneau, Ed., Plenum Press: New York; pp. 172-199,
1973.
22. Dedrick, R. L., Meyers, C. E., Bungay, P. M. and De Vita, V. T.:
Pharmacokinetic rationale for peritoneal drug administration in the
treatment of ovarian cancer. Cancer Treatment Rep. 62: 1—11, 1978.
23. Lee, I. P. and Dixon, R. L.: Factors influencing reproduction and genetic
toxic effects on male gonads. Environ. Health Perspect. 24: 117—127,
1978.
24. Beach, F. A.: Animal models for human sexuality. In: CIBA Foundation
Symposium, No. 62. Sex, Hormones, and Behavior, Mar. 14-16, 1978,
London, England, Elsevier-North Holland: New York; pp. 133-144, 1979.
25.McClintock, M. K. and Adler, N. T.: The role of the female during
copulation in wild and domestic rats (Rattus norvegicus). Behavior 67:
67-96, 1978.
26. Madlafousek, J. and Hlinak, Z.: Sexual behavior of the female laboratory
rat: inventory, patterning, and measurement. Behavior 63: 129—174,
1977.
27. Dewsbury, D. A.: A quantitative description of the behavior of rats during
copulation. Behaviour 29: 154-178, 1967.
28. Dewsbury, D. A.: Description of sexual behavior in research on hormone-
behavior interactions. In: Endocrine Control of Sexual Behavior, C. Beyer,
Ed., New York: Raven Press, pp. 3-32,1979.
29.McClintock, M. K., Toner, J. P., Adler, N. T. and Anisko, J. J.: Postejacu-
latory quiescence in female and male rats: consequences for sperm
transport during group mating. J. Comp. Physiol. Psycho!.; in press, 1982.
30. Beach, F. A. and Jordan, L.: Sexual exhaustion and recovery in the male rat.
Q. J. Exp. Psychol. 8: 121-133,1956.
Sl.Larsson, K.: Conditioning and Sexual Behavior in the Male Albino Rat.
Stockholm: Almquist & Wiksell; 1956.
32. Sachs, B. D. and Barfield, R.: Functional analysis of masculine copulatory
behavior in the rat. Adv. Study Behav. 7: 91-154, 1976.
-------
33
34
35
36
37
38.
39.
40.
41.
42.
43.
44
45.
46.
47.
48.
49.
143
.Young, W. C.: Psychobiology of sexual behavior in the guinea pig. Adv.
Study Behav. 2: 1-110,1969.
Young, W. C. and Grunt, J. A.: The pattern and measurement of sexual
behavior in the male guinea pig. J. Comp. Physiol. Psychol. 44: 492-500
1951.
Dewsbury, D. A.: Copulatory behavior of rats: Variations within the dark
phase of the diurnal cycle. Commun. Behav. Biol. 1: 373-377, 1968.
Fadem, B. H., Barfield, R. J. and Whalen, R.E.: Dose-response and
time-response relationships between progesterone and the display of
patterns of receptive and preceptive behavior in the female rat. Horm.
Behav. 13: 40-48, 1979.
Goy, R. W. and Young, W. C.: Strain differences in the behavioral responses
of female guinea pigs to alpha-estradiol benzoate and progesterone.
Behaviour 10: 340-354, 1957.
Walker, W. A. and Feder, H. H.: The comparative potency of various steroids
to complete the priming process for lordosis in guinea pigs. Horm. Behav
12: 299-308, 1979.
Diakow, C. and Dewsbury, D. A.: A compuative description of the mating
behavior of female rodents. Anim. Behav. 26: 1091-1097,1978.
McClintock, M. K., Anisko, J.J. and Adler, N.T.: Group mating among
Norway rats. I. Sex differences' in the pattern and neuroendocrine
consequences of copulation. Anim. Behav.; in press, 1982.
. Beach, F. A.: Cross-species comparisons and the human heritage. In: Human
Sexuality in Four Perspectives, F. A. Beach, Ed., Johns Hopkins University
Press: Baltimore; pp., 296-316, 1977.
.Beach, F. A.: Animal models and psychological inference. In: Human
Sexuality: A Developmental and Comparative Approach, H. A.
Katchadourian, Ed., University of California Press: Berkeley ;pp 98-112
1979.
. Adams, D., Gold, A. and Burt A.: Rise in female-initiated sexual activity at
ovulation and its suppression by oral contraceptives. N Engl J Med 299-
1145-1150,1978.
Persky, H., Lief, H., Strauss, D., Miller, W. and O'Brien, C.D.: Plasma
testosterone level and sexual behavior of couples. Arch. Sex. Behav. 7:
157-173, 1978.
Byrne, D. and Scheffield, J.: Response to sexually arousing stimuli as a
function of repressing and sensitizing defenses. J. Abnorm. Psychol 70:
114-118,1965.
Byrne, D., Fisher, J. D., Lamberth, J. and Mitchell, H. E.: Evaluations of
erotica: Facts or feelings. J. Pers. Soc. Psychol. 29: 111-116, 1974.
Kaats, A. and Davis, K.: Effects of volunteer biases in studies of sexual
behavior and attitudes. J. Sex. Res. 7: 219-227, 1971.
Bauman, K.: Volunteer bias in a study of sexual knowledge, attitudes, and
behavior. J. Marr. Fam. 35: 27-31,1975.
Kaplan, H. S.: The New Sex Therapy, Brunner/Mazel: New York; 544 pp
1974. '
-------
144
50. Abel, E. L.: A review of alcohol's effects on sex and reproduction. Drug
Alcohol Depend. 5: 321-332,1980.
51. Barnes, T. R., Bamber, R. W. and Watson, J.P.: Psychotropic drugs and
sexual behaviour. Br. J. Hosp. Med. 21: 594-600,1979.
52. Clemens, L. G., Popham, T. V. and Ruppert, P. H.: Neonatal treatment of
hamsters with barbiturate alters adult sexual behavior. Dev. Psychobiol.
12: 49-59,1979.
SS.Dalterio, S. L.: Perinatal or adult exposure to cannabinoids alters male
reproductive functions in mice. Pharmacol. Biochem. Behav. 12:
143-153,1980.
54.Dalterio, S. and Bartke, A.: Perinatal exposure to cannabinoids alters male
reproductive function in mice. Science 205: 1420-1422, 1979.
55. Gordon, J. H., Bromley, B. L., Gorski, R.A. and Zimmermann.E.:
Delta9-tetrahydrocannabinol enhancement of lordosis behavior in estrogen
treated female rats. Pharmacol. Biochem. Behav. 8: 603-608, 1978.
56. Kostellow, A. B., Ziegler, D., Kunar, J., Fujimoto, G. I. and Morrill, G. A.:
Effect of cannabinoids on estrous cycle, ovulation, and reproductive
capacity of female A/J mice. Pharmacology 21: 68-75, 1980.
57. Kumar, R., Mumford, L. and Teixeira, A. R.: Sexual behavior in morphine-
dependent rats [proceedings]. Br. J. Pharmacol. 62: 389P-390P, 1978.
58. Mumford, L. and Kumar, R.: Sexual behaviour of morphine-dependent and
abstinent male rats. Psychopharmacologie (Berlin) 65: 179-185, 1979.
59. Ostrowski, N. L., Stapleton, J. M., Noble, R. G. and Reid, L. D.: Morphine
and naloxone's effects on sexual behavior of the female golden hamster.
Pharmacol. Biochem. Behav. 11: 673-681, 1979.
60. Rosenkrantz, H.: Effects on cannabis on fetal development of rodents. Adv.
Biosci. 22-23: 479-499, 1978.
61.Vorhees, C. V., Brunner, R. L. and Butcher, R. E.: Psychotropic drugs as
behavioral teratogens. Science 205: 1220-1225, 1979.
62.Olusi, S. 0., Oke, O. L. and Odusote, A.: Effects of cyanogenic agents on
reproduction and neonatal development in rats. Biol. Neonate 36:
233-243,1979.
63. Palmer, A. K., Cozens, D. D., Spicer, E. J. and Worden, A.N.: Effects of
lindane upon reproductive function in a three-generation study in rats.
Toxicology 10: 45-54,1978.
-------
STEERING COMMITTEE
K. Diane Courtney, Ph.D.
Research Pharmacologist
Pesticides and Toxic Substances
Effects Laboratory
U.S. Environmental Protection
Agency
Research Triangle Park, NC 27711
Wayne M. Galbraith, Ph.D.f
Toxicologist
Office of Research and Development
U.S. Environmental Protection
Agency
Washington, D.C. 20460
(EPA Co-Project Officer)
Richard M. Hoar, Ph.D.
Head of Teratology and Assistant
Director of Toxicology
Department of Toxicology
Hoffmann—LaRoche, Inc.
Nutley, NJ07110
(Chairman of Reproduction
Groups)
E. Marshall Johnson, Ph.D.
Professor and Chairman
Department of Anatomy
Director, Daniel Baugh Institute
of Anatomy
Jefferson Medical College
Thomas Jefferson University
Philadelphia, PA 19107
(Chairman of Developmental Group)
Robert M. Pratt, Ph.D.*
Chief, Experimental Teratogenesis
Section
Laboratory of Reproductive &
Developmental Toxicology
National Institute of Environmental
Health Sciences
National Institutes of Health
Research Triangle Park, NC 27709
Michael G. Ryon, M.S.
Information Analyst
Chemical Effects Information
Center
Oak Ridge National Laboratory
Oak Ridge, TN 37830
Peter Voytek, Ph.D.
Director, Reproductive Effects
Assessment Group
Office of Research and
Development
U.S. Environmental Protection
Agency
Washington, D.C. 20460
(EPA Co-Project Officer)
*Attended only the St. Louis workshop.
•(•Currently, Acting Chief of the Toxicology Branch, Research Division, Chemi-
cal Systems Laboratory, Aberdeen Proving Ground, Maryland 21001.
145
-------
-------
PARTICIPANTS
Rupert P. Amann, Ph.D.
Professor of Physiology and
Biophysics
Animal Reproduction Laboratory
Colorado State University
Fort Collins, CO 80525
J. Michael Bedford, Ph.D.
Professor of Obstetrics and
Gynecology and of Anatomy
Cornell Medical School
New York, NY 10021
(Chairman of Male Reproduction
Group)
Allan R. Beaudoin, Ph.D.
Assistant Chairman and Professor
Department of Anatomy
University of Michigan Medical
School
Ann Arbor, MI 48104
Kenneth B. Bischoff, Ph.D.
Chairman, Department of Chemical
Engineering
Professor of Biomedical and
Chemical Engineering
University of Delaware
Newark, DEI 9711
William J. Bremner, M.D., Ph.D.
Chief, Endocrinology Section
Veterans Administration Medical
Center
Associate Professor of Medicine
and of Obstetrics and Gynecology
University of Washington School
of Medicine
Seattle, WA 98108
Charles C. Brown, Ph.D.*
Statistician, Biometry Branch
National Cancer Institute
Bethesda, MD 20205
Mildred S. Christian, Ph.D.
Director of Research
Argus Research Laboratories
Perkasie, PA 18944
James H. Clark, Ph.D.
Professor of Cell Biology
Baylor College of Medicine
Houston, TX 77030
(Chairman of Female
Reproduction Group)
Thomas F. X. Collins, Ph.D.
Chief, Mammalian Reproduction
and Teratology
Division of Toxicology
Bureau of Foods
Food and Drug Administration
Washington, D.C. 20204
Donald A. Dewsbury, Ph.D.*
Professor of Psychology
University of Florida
Gainesville, FL 32611
Larry L. Ewing, Ph.D.*
Professor of Reproduction
Biology
School of Hygiene and Public
Health
Johns Hopkins University
Baltimore, MD 21205
*Attended only the St. Louis workshop.
147
-------
148
Robert H. Foote, Ph.D.
Professor of Animal Science
Cornell University
Ithaca, NY 14850
David W. Gaylor, Ph.D.
Director, Division of Biometry
National Center for Toxicological
Research
Jefferson, AR 72079
Arnold A. Gerall, Ph.D.*
Professor of Psychology
Tulane University
New Orleans, LA 70118
Robert W. Goy.Ph.D.*
Director, Wisconsin Regional
Primate Research Center
University of Wisconsin
Madison, WI 53706
(Chairman of Behavior Group)
Arthur F. Haney, M.D.
Assistant Professor of Obstetrics
and Gynecology
Duke University Medical Center
Durham, NC 27710
W. LeRoy Heinrichs, Ph.D., M.D.*
Chairman and Professor
Department of Gynecology and
Obstetrics
Stanford University
Stanford, CA 94305
Mary C. Henry, Ph.D.*
Research Pharmacologist
Environmental Protection Research
Division
U.S. Army Biomedical Engineering
Research and Development
Laboratory
Fort Detrick
Frederick, MD 21701
Jerry Highfffl, M.S.
Statistician, Health Effects
Research Laboratory
U.S. Environmental Protection
Agency
Research Triangle Park, NC 27711
Carol J. Hogue, Ph.D.
Associate Professor of Biometry
University of Arkansas for
Medical Sciences
Little Rock, AR 72201
Donald E. Hutchings, Ph.D.
Research Scientist
Department of Behavioral
Physiology
New York State Psychiatric
Institute
Assistant Professor of Medical
Psychology
Department of Pediatrics
Columbia University
New York, NY 10031
Harold Kalter, Ph.D.
Research Associate
Children's Hospital Research
Foundation
Professor of Research
Department of Pediatrics
University of Cincinnati
Cincinnati, OH 45229
*Attended only the St. Louis workshop.
-------
149
Carole A. Kimmel, Ph.D.*
Chief, Perinatal and Postnatal
Evaluation Branch
Research Pharmacologist
Division of Teratogenesis
Research
National Center for Toxicological
Research
Jefferson, AR 72079
Devendra M. Kochhar, Ph.D.
Professor of Anatomy
Jefferson Medical College
Thomas Jefferson University
Philadelphia, PA 19107
Donna Kuroda, Ph.D.
Physical Sciences Administrator
Reproductive Effects
Assessment Group
U.S. Environmental Protection
Agency
Washington, D.C. 20460
Donald R. Mattison, M.D.
Medical Officer
Pregnancy Research Branch
National Institute of Child
Health and Human Development
National Institutes of Health
Bethesda, MD 20205
Martha K. McClintock, Ph.D.*
Assistant Professor of Behavioral
Sciences
University of Chicago
Chicago, IL 60637
Wilbur P. McNulty, M.D.
Chairman, Laboratory of Pathology
Oregon Regional Primate Research
Center
Beaverton, OR 97006
Marvin L. Meistrich, Ph.D.
Associate Professor of
Experimental Radiotherapy
University of Texas System
Cancer Center
M.D. Anderson Hospital Tumor
Institute
Houston, TX 77030
Eugene F. Oakberg, Ph.D.*
Senior Research Staff Member
Mammalian Genetics and
Development Section
Biology Division
Oak Ridge National Laboratory
Oak Ridge, TN 37830
James W. Overstreet, Ph.D., M.D.*
Associate Professor of Human
Anatomy and of Obstetrics and
Gynecology
University of California, Davis
Medical School
Davis, CA 95616
John C. Porter, Ph.D.
Professor of Physiology and of
Obstetrics and Gynecology
University of Texas Health Science
Center at Dallas
Southwestern Medical School
Dallas, TX 75235
Lynn Rosenberg, Sc.D.
Assistant Research Professor
Boston University
Biostatistician and
Epidemiologist
Drug Epidemiology Unit
Boston University Medical
Center
Cambridge, MA 02138
*Attended only the St. Louis workshop.
-------
150
Griff T. Ross, Ph.D.,M.D.f
Deputy Director of the
Clinical Center
National Institutes of Health
Bethesda, MD 20205
Liane B. Russell, Ph.D.f
Head of Mammalian Genetics and
Teratology Section
Biology Division
Oak Ridge National Laboratory
Oak Ridge, TN 37830
Carol Sakai, Ph.D.
Reproductive lexicologist
Reproductive Effects Assessment
Group
Office of Research and
Development
U.S. Environmental Protection
Agency
Washington, D.C. 20460
Thomas H. Shepard, M.D.*
Professor of Pediatrics
Head, Central Laboratory for
Human Embryology
University of Washington
Seattle, WA 98195
Richard G. Skalko, Ph.D.
Professor and Chairman
Department of Anatomy
College of Medicine
East Tennessee State University
Johnson City, TN 37614
Kate Smith, Ph.D.*
Developmental Toxicologist
Health Effects Research Laboratory
U.S. Environmental Protection
Agency
Cincinnati, OH 45268
Robert E. Staples, Ph.D.
Staff Teratologist
Haskell Laboratory (DuPont)
Wilmington, DE 19898
Robert G. Tardiff, Ph.D.f
Executive Director, Board on
Toxicology and Environmental
Health Hazards
National Academy of Sciences
Washington, D.C. 20418
William J. Waddell, M.D.
Professor and Chairman
Department of Pharmacology and
Toxicology
School of Medicine
University of Louisville
Louisville, KY 40292
Ronald J. Young, Ph.D.
Associate Professor and Research
Associate
Department of Obstetrics and
Gynecology
Cornell University Medical College
New York, NY 10021
* Attended only the St. Louis workshop.
t Attended only the Atlanta workshop.
-------
REVIEWERS
The assistance of the following persons in the review process is gratefully
acknowledged. The final content of the report is the responsibility of the steer-
ing committee and the group chairmen.
Aaron Blair, Ph.D.
Environmental Epidemiology
Branch
National Institutes of Health
Joseph Borzelleca, Ph.D.
Department of Pharmacology
Medical College of Virginia
Robert Dedrick, Ph.D.
Biomedical Engineering and
Instrumentation Branch
National Institutes of Health
James Emerson, Ph.D.
Life Sciences
Coca Cola Company
Michael Farrow, Ph.D.
Genetic Toxicology Department
Hazleton Laboratories America,
Inc.
Ernst Freese, Ph.D.
Laboratory of Molecular Biology
National Institutes of Health
Vera Glocklin, Ph.D.
Bureau of Drugs
U.S. Food and Drug Administration
Andrew G. Hendrickx, Ph.D.
California Primate Research
Center
University of California, Davis
Kundan S. Khera, Ph.D.
Health Protection Branch
Health and Welfare Canada
Renate Kimbrough, M.D.
Toxicology Branch
Centers for Disease Control
George Levinskas, Ph.D.
Environmental Assessment and
Toxicology
Monsanto Company
Lawrence B. Mellett, Ph.D.
Scientific Liaison and Compliance
Revlon Health Care Group
Roy Mundy, Ph.D.
Department of Pharmacology
University of Alabama,
Birmingham
Frederick Oehme, D.V.M., Ph.D.
Comparative Toxicology Laboratory
Kansas State University
Anthony K. Palmer
Huntingdon Research Centre
Bobby Joe Payne, D.V.M., Ph.D.
Director of Pathology
Toxicity Research Laboratories,
Ltd.
Harold M. Peck, M.D.
Safety Assessment
Merck Institute for Therapeutic
Research
David Rail, M.D., Ph.D.
National Institute of Environmental
and Health Sciences
151
-------
152
Robert Scala, Ph.D.
Medical and Environmental Health
Department
Exxon Corporation
Charlotte Schneyer, Ph.D.
Laboratory of Exocrine Physiology
University of Alabama,
Birmingham
Bernard A. Schwetz, D.V.M., Ph.D.
Health and Environmental
Sciences
Dow Chemical U.S.A.
Marshall Steinberg, Ph.D.
Life Sciences Division
Hazleton Laboratories America, Inc.
Clarence J. Terhaar, Ph.D.
Toxicology Section
Eastman Kodak Company
Hanspeter Witschi, M.D.
Biology Division
Oak Ridge National Laboratory
Gerhard Zbindin, M.D.
Institute of Toxicology
-------
INDEX
abstinence interval, 60, 82
acceptable daily intake (ADI), 9,
48,56, 110
acceptable dosage level, 48, 111,
112
accessory sex glands, 6,47 (Table 8),
49 (Table 9), 50 (Table 10), 51,
60,73,75-77,79,84
acute, 3,57, 70, 71,99, 122
adenosis, 8 (Table 1), 37
adolescence, 126
adrenal, 8 (Table 1), 127
adult toxic dose, 108, 111,112
age, 7, 10,11, 14 (Table 2), 16,
19-21, 28, 48 (Table 8), 50
(Table 10), 52, 53, 60, 74, 75,
82, 111, 118
amenorrhea, 20, 37, 39
androgen, 5, 6,14, 19, 20, 30, 33
(Table 6), 37, 62, 73, 93,126,
128-130
animal
model, 11, 17, 18,20,23,28,
30,32, 41-43,44 (Table 7),
50 (Table 10), 52, 56, 57, 58
(Table 11), 60, 62, 69, 71,83,
110 (Table 12), 122, 124,
128-130, 140
testing, 2, 6, 13, 19, 22,25-27,
42, 46, 47 (Table 8), 49
(Table 9), 50 (Table 10), 52,
55-57, 58 (Table 11), 59-61,
81,85,86,100,101, 104,106,
108, 109, 110(Table 12), 111,
112,119, 130-137, 139, 140
anovulation, 8 (Table 1), 17, 20, 37
artificial insemination, 43, 44
(Table 7), 58 (Table 11), 85
attractivity, 124, 125, 131, 134,
135,140
azoospermia, 54, 59
bioaccumulation, 48, 57, 58
(Table 11), 59, 102, 108
breast, 8 (Table 1)
case-control study, 118-120
cauda epididymidis, 44 (Table 7),
49,51,59,72,74,83
cell culture, 22-24, 31, 32, 33
(Table 6), 62, 109, 110
(Table 12)
chronic, 8 (Table 1), 46, 48
(TableS), 51,70, 71,88,99,
111, 122
cleft lip, 101
cleft palate, 101
coefficient of variation (CV), 46,
50 (Table 10), 60, 62, 75, 77, 81
cohort-studies, 118, 119, 121, 137
computerized integrated data base,
4 (Fig. 1), 5-7, 9, 10, 121
conception, 43, 118
conceptive ability, 6, 17, 31
conceptus, 1, 99, 100, 103, 104,
106-110, 112, 122
congenital defects (or malforma-
tions), 100, 118
control groups, 16, 46, 53-57, 61,
70-72, 74, 77, 78, 80-84, 123,
132, 137-139
copulation plug, 16
copulatory behavior, 17,132-134
153
-------
corpus luteum (lutea), 17,30, 32,
37,47 (Table 8), 50 (Table 10),
83,85
correlational studies, 118
cost/benefit, 16, 17, 49 (Table 9),
108,139
critical periods of development,
20,102, 103
decidual implantation, 104, 105
demographic studies, 3, 118
developmental abnormalities (or
defects or malformations), 99,
100, 103-107,110 (Table 12),
118,126,129,130
developmental toxic dose, 108,
110(Table 12), 111, 112
diestrus, 20,37
DNA, 31, 49 (Table 9), 62, 87
dog, 43, 44 (Table 7), 79, 131
domestic animals, 32, 42, 43,46,
70, 74, 76, 79, 107, 108
dopamine, 11, 21, 24,37
dose, 6, 16, 18-20,22,26,27,
47 (Table 8), 51,52, 55-57,
59,61,83,87,88,95,101-104,
106-109,111,121,122, 130,
134, 136
dose-response (relationship), 2, 9,
18, 20, 23, 24, 26-28,47 (Table 8),
48, 57, 79, 85, 87, 88,101,103,
109-111, 112 (Table 12), 119
ductus deferens, 72
ejaculate, 41-43, 46,47 (Table 10),
49, 50 (Table 10), 51, 54, 55, 59,
60, 69, 73-79, 82, 84, 85, 87, 94,
133
ejaculation, 42, 43, 44 (Table 7),
51,60,74,76,77,94, 132,133
emoryo, 47 (Table 8), 50 (Table 10),
59,83,85,86, 101-103,105,
109, 110 (Table 12), 112
154
embryonic development, 41, 62,
102-105
emission, 95
endocrine, 1, 26,43, 46,47
(Table 8), 50 (Table 10), 53, 79,
80, 103, 105, 107, 126, 134
epidemiology (epidemiologic), 1,
3,5,12,53,56, 106,117,119-
121, 136-138, 140
epididymis, 41, 46,47 (Table 8),
49 (Table 9), 50 (Table 10), 51,
59, 70, 72-75, 78, 83, 84
estradiol, 14, 18, 19, 22, 26, 33
(Table 6), 37, 127, 129, 134,
136
estrogen, 5, 9, 11, 13, 14 (Table 2),
18,19,21,23,25-27,30,33
(Table 6), 37, 38, 126-129, 136,
138
estrogenicity, 5, 6, 9, 11, 13, 14,
18, 19, 21-23, 27
estrus, 14, 15 (Table 3), 16, 17,
20, 38, 104, 133, 134, 136
extrapolation, 10-12,18-20, 30,
32,56, 100, 102, 104, 111, 112,
117,122, 123, 130, 136
fallopian tubes, 8 (Table 1)
false negative, 16, 17, 109
false positive, 5, 17, 109
fecundity, 17
fertility, 16, 17, 28,4143,44
(Table 7), 46,47 (Table 8), 50
(Table 10), 51,53-56, 59, 60,
78,83-87, 107,118,120
fertilization, 41,43, 62
fetal development, 10, 14, 20, 62,
102-105, 129
fetus, 2, 5, 102, 105, 106, 122,
129, 130
follicle, 11,21, 28, 29 (Table 4),
30, 32, 33 (Table 6), 38,135
follicle-stimulating hormone (FSH),
21, 30, 38, 41,47 (Table 8), 50
(Table 10), 53,54, 59,61,62,
79-84, 95
-------
gamete, 8 (Table 1), 27, 28, 104
germ cell, 49 (Table 9), 71, 74
germinal epithelium, 49, 54, 57,
59,69,71,73,78
gestation, 8 (Table 1), 18, 20,
102-104
gonad, 30, 54, 104, 122, 124, 126,
128, 136
gonadotropin, 10, 17, 21, 22, 26,
30-32, 38, 136
gonadotropin-releasing hormone
(GnRH), 21, 22, 24, 47 (Table 8),
50 (Table 10), 54, 61, 80, 82
granulosa, 30, 32, 33 (Table 6)
gross anatomical (or structural)
defects, 106,107
guinea pig, 17, 27, 127-129, 131,
133-135, 140
hamster, 43, 47 (Table 8), 82, 84,
86, 87, 127, 129
hazard, 1,6,41-43, 53, 62, 69, 85,
99, 100, 106, 108, 109, 111,
112, 117-121
heterospermic insemination, 62
hormone, 5, 6, 8 (Table 1), 9,11,
21,25,27, 30, 33 (Table 6),
41, 44 (Table 7), 47 (Table 8),
50 (Table 10), 53,54, 59,61,
62,79-84,95,124,126-131,
133, 136
human, 1-3, 5-7, 9-12, 16-20, 22,
24, 27, 28, 30-32, 41-43, 44
(Table 7), 47 (Table 8), 49
(Table 9), 50 (Table 10), 52-62,
70, 75, 76, 79-82, 84-88, 99-
102, 106-112, 117, 119,120,
122-130, 134-140
human chorionic gonadotropin
(HCG),31
hypothalamus (hypothalamic), 5,
8 (Table 1), 11, 20,21,24-27,
38,80, 128, 131
hypothesis-generating studies,
117-119
155
implantation, 14, 104
infertility, 1, 5, 42, 43, 52, 54, 55,
80,82,118
intromission, 132, 133
inutero, 104, 106
in vitro, 10, 18, 22-24, 30, 32, 33
(Table 6), 44 (Table 7), 47
(Table 8), 61,62, 79,82,84,
85, 109, 110 (Table 12)
in vivo, 10, 14, 18,19, 21-23, 30,
31,57,62,79,85,86,108, 109
lactation, 6, 8 (Table 1), 14, 16,
18,38,104,105
leydigcell, 61,62, 72
libido, 20,43, 84, 124,127, 132,
140
litter size, 18, 84
live birth index, 18
lordosis, 26, 27, 135
luteinizing hormone (LH), 21, 22,
25, 30, 38, 41,47 (Table 8),
50 (Table 10), 54,59,61,62,
79-84, 95
masculinization, 5,10, 20,40,
126, 129
maternal-fetal exchange, 102
maternal toxicity, 110
mating, 14, 15 (Table 3), 16, 17,
26, 27, 44 (Table 7), 46, 58
(Table 11), 59, 62,83, 85,86,
104, 133
mating behavior, 14, 20, 25, 26
maximum tolerated dose (MTD),
16,47 (Table 8), 49 (Table 9),
57, 83, 88, 95
menopause, 8 (Table 1), 28, 38
menstruation, 17; 30, 138
model system, 9-11, 28, 32, 43,
60,69,71, 101, 110 (Table 12),
122, 124, 130
motility (or motile), 42,43, 46,
47 (Table 8), 50 (Table 10), 51,
53, 60, 72, 73, 75, 77-79, 82-84,
95
-------
mounts, 124, 126, 132, 133, 135
mouse (mice), 6, 11, 28, 43,44
(Table?), 51,61,74, 102, 110
(Table 12), 127,129
multigenerational, 6, 15 (Table 3),
104, 105
nonmotile spermatozoan, 95
no-observed-effect level (NOEL)
or (no-adverse-effect level), 56,
103, 110
oligomenorrhea, 20, 38
oligozoospermia, 54
oocyte, 8 (Table 1), 11, 18,27,
28, 29 (Table 4), 39, 84
oogenesis, 28
oral contraceptive, 11
organogenesis, 101, 102, 105,
106,122
ovary, 8 (Table 1), 16-18, 21, 25,
27, 28, 30-32, 33 (Table 6),
127, 134-136
ovotoxicity, 11
ovulation, 10, 14, 17,20, 21, 39
parturition, 8 (Table 1), 39
pharmacokinetics, 1,56, 101,
103, 111, 112, 117, 121-123,
130
phenotypic transformation, 20
pituitary, 8 (Table 1), 10, 21-26,
54, 80, 128
placenta, 8 (Table 1), 102, 103,
112, 129,130
placental transfer, 8 (Table 1), 102
population-based registries, 118
postpartum, 14, 15 (Table 3), 16
potency, 18, 19,22, 23, 128
pregnancy, 5, 6, 8 (Table 1), 16,
20,47 (Table 8), 50 (Table 10),
52,105,110,117,118, 120
156
primate, 10, 17, 22, 23, 30,43,
44 (Table 7), 74, 131
proceptivity, 135
progestagen, 26, 136
progesterone, 11,21, 26, 27, 30,
33 (Table 6), 39, 136
prolactin, 8 (Table 1), 21, 23, 24,
39
prostaglandin, 39
puberty, 6, 8 (Table 1), 14, 39, 75,
49 (Table 9)
qualitative, 22, 43, 74, 87
reproductive toxicity screen, 3,
4 (Fig. 1), 6, 7,9, 13, 18, 19
risk assessment, 46, 100, 128,
140
test, 10, 71
Quantitative, 21, 25, 28, 43, 74,
87, 121, 128
reproductive toxicity test, 3,4
(Fig. 1), 7, 10, 18, 24, 55, 71
risk, 4, 5,32,46, 56,108, 110,
140
rabbit, 43,44 (Table 7), 46,47
(Table 8), 49 (Table 9), 50
(Table 10), 51, 56, 57, 58
(Table 11), 59, 60,62, 70, 74,
76,77,79,81,83,85,102,105
rat, 13, 14 (Table 2), 17,19, 20,
22, 23, 25, 26, 32, 43, 44
(Table 7), 47 (Table 8), 49
(Table 9), 50 (Table 10), 51,
56, 57, 58 (Table 11), 59-62,
73, 74, 79, 81, 83, 85, 102,
105, 110 (Table 12), 124, 127-
129, 131, 133-135, 140
receptivity, 9, 124, 126, 127,132,
134-136
registries of birth defects, 120
radioimmunoassay, 22,23, 31,
81,82, 130
-------
reproductive capabilities, 6, 8
(Table 1)
reproductive capacity, 62
reproductive dysfunction, 49
(Table 9), 53, 55, 82
reproductive function, 20, 41,
42, 53, 56-58, 137
reproductive performance, 14, 55
reproductive toxicants, 3, 6, 7, 9-
11,48,52,54,80,81,86,123,
137
research, 2, 9-12, 49 (Table 9),
60-62,76, 112, 120, 121, 123,
139, 140
reversibility, 2, 7,26,47 (Table 8),
59,81,126, 139
risk assessment, 1, 3, 4 (Fig. 1), 7,
11-13,42,52,56,57,61,69,
86,99, 117, 121, 123,139
risk estimation, 2, 99, 100, 109,
111,112
rodent, 17, 18, 20, 27
route of administration (or
exposure), 9, 16, 26,27, 57, 83,
101, 121, 130
safety factor, 9, 52, 56, 57, 110-
112
screening system (or procedures),
3, 4 (Fig. 1), 5-7, 14 (Table 2),
16-18, 25, 27, 46, 47 (Table 8),
50 (Table 10), 53,54, 57, 61,
62,71,74,78,79,87,107,109,
110 (Table 12), 124, 129, 131,
137
scrotol circumference, 69, 70
semen, 41-43, 44 (Table 7), 46,
47 (Table 8), 49, 50 (Table 10),
51, 53-55, 58 (Table 11), 59-62,
69, 73-79,82, 85-87, 96
seminal
characteristics, 60, 70, 74, 75
collections, 46, 74, 76, 82
157
fluid (or plasma), 44 (Table 7),
47 (Table 8), 49 (Table 9), 50
(Table 10), 53,54,59-61,75,
76, 79, 80, 82
volume, 47 (Table 8), 50 (Table 10),
60, 74, 75, 82, 84, 96
seminiferous epithelium, 44 (Table 7),
47 (Table 8), 48, 51,57, 71,74,
75,83,84,88,93,94,96
Sertoli cell, 62, 71,72,83,84
sexual
behavior, 1, 8 (Table 1), 10, 20,
25-27,43,62,80,117,123-131,
134-140
dysfunction, 53, 123, 125., 137-139
function, 26, 80, 104,124,137,
138, 140
gratification, 125, 138-140
initative, 124-126, 131,138
responsiveness, 124-126, 131,
135,136
short-term test, 31,62, 107-109,
110 (Table 12)
sonography, 53
species, 7, 9, 11, 16, 17, 28, 30-32,
43, 46, 49 (Table 9), 51,56,
57,61,69-71,74-76,79-81,
100-106, 110-112, 117, 122-124,
126, 128-131, 134, 140
sperm, 16, 41-43, 44 (Table 7),
46, 47 (Table 8), 49 (Table 9),
50 (Table 10), 51,54, 59, 69,
62,69, 71, 72, 74-76, 78-80,
82, 84-87, 132
abnormalities, 60-62, 78, 85, 87
chromatin, 62
concentration, 41, 42, 46, 47
(Table 8), 50 (Table 10), 72,
76, 77, 82
count, 42, 53, 54, 72, 76, 138
genone, 41,62,85, 86
head, 55, 62, 73, 78, 83
maturation, 43
morphology, 41, 42, 46, 47
(Table 8), 50 (Table 10), 51,
53, 55, 59-62, 73, 75, 78, 79,
84,87
-------
motility, 42,43,46,47 (Table 8),
50 (Table 10), 51,53,60,72,
73, 75, 77-79, 82-84, 93, 95
number, 41,46,47 (Table 8), 50
(Table 10), 53, 55,72, 76-79,
82, 84, 85, 98
production, 42,44 (Table 7), 51,
54, 60, 70, 75-77, 80, 85, 94 ,
size, 55
transport, 44 (Table 7), 47
(Table 8), 59, 62, 70, 85
spermatid, 44 (Table 7), 47
(Table 8), 50 (Table 10), 51,
58,70,71,83,84
spermatocytes, 44 (Table 7), 47
(Table 8), 49, 51,71,72, 83, 84
spermatogenesis, 43, 70, 71, 74,
78, 80, 86, 94, 97
spermatogonia, 42,44 (Table 7),
49,51,59,71
spermatozoa, 42,46,49 (Table 9),
55, 60, 62, 69, 70, 72-74, 78,
83, 85-87,94-98
spermatozoal concentration, 42,
76,84
spermicidal, 79
spermiogenesis, 51
spontaneous luteinization, 32
sterility, 71, 80,82,86
steroid, 16,17,21,28,30,31
(Table 5), 32, 33 (Table 6),
•127,136,137
steroidogenesis, 8 (Table 1), 10,
16,21,28-30, 31 (Table 5), 32,
33 (Table 6), 39,136,137
structure-function, 3-7, 9, 10, 56 .
subchronic, 3,48, 57
surveillance study, 52, 54,120,
121, 136, 137
158
testis, 42,44 (Table 7), 47
(Table 8), 49 (Table 9), 50
(Table 10), 51,54, 56, 69,70,
71,73,74,79,80,83-85, 127,
128, 131
testis weight, 44 (Table 7), 47
(TableS), 50 (Table 10), 58
(Table 11), 70, 77, 83, 84
testosterone (T), 6, 20, 30, 33
(Table 6), 39, 41,47 (Table 8),
48, 50 (Table 10), 54, 59, 61,
62,69,79-84,97,128, 129,131
thecol, 30, 32, 33 (Table 6)
threshold level, 2, 42, 56, 88, 103,
110, 111, 130
thyrotropin, 23
tonometry, 49 (Table 9), 53, 60
uncertainty factor, 56
uterus (or uterine), 5, 8 (Table 1),
13, 14 (Table 2), 18, 19, 112,
122
vagina, 8 (Table 1), 43, 59, 74, 84,
132, 133, 136, 138
vaginal
opening, 5, 13,14 (Table 2), 18,
20
smear, 16, 20, 58 (Table 11),
135
validation, 7, 10, 16, 32, 46, 49
(Table 9), 55, 84, 85, 99, 108,
109,112,119, 120
vesicular gland, 47 (Table 8), 50
(Table 10), 73, 78, 83
videotape (video), 46,47 (Table 8),
48, 50 (Table 10), 60, 73, 78, 82
virilization, 8 (Table 1), 40, 126
testicular
function, 41,42, 46, 52-56,58,
60,61,69,71,72,74,78-80,
82
size, 44 (Table 7), 47 (Table 8),
50 (Table 10), 51,53, 54, 60,
69, 70, 74,84
zona pellucida, 84, 86
zygote, 8 (Table 1), 104
US. GOVERNMENT PRINTING OFBCE: 1984 - 759-102/10704
-------
-------
------- |